orce as a
approach that was simpler to learn and use th MO structured tt
an RT (Stephenson ,,
Other types of structured trees such as !JSfe fl h 1991).
, – , ms owe arts, can be us
a syseem s structure graphically. For accident • . . ed to sho,,
mvesugauons an i .
uses the tree to trace back through the system d find ‘ nvesngaror
an faults .
The Fault Tree Approach
The fuse seep in constructing a fault tree is to determ· h
. . _ . 1ne t e top evem.
For acadent mvesnganons, the top event is the accident · · d , tnJury, or amage
that occurred (Hammer 1993) . Events that had to happen in order for the
accident to happen are listed on the next tier of the tree. Causal factors-
fixable siruations or correctable areas-are on the bottom tier of the tree.
The corrective actions the accident investigator recommends will be geared
to fixing these problems.
Symbols and gates are part of the fault-tree diagram. The mosr common
tree symbols and gates are illustrated in Exhibit 10.3. Since the purpose of
this chapter is to describe how trees can be used in accident investigations co
Chapter 10: Tree A nalysis
fault tree analysis methodologies and symbols that
al factors, di d . e ,nos, _ – vestigation are not scusse .
t1ettf111u1 ‘ to ncCident 1Il . –
0, ,ppl) . ediate” events m fault-tree termmology. Failures Jo fl re “1nterm . .
,vents a . all diagramed under tntermediate events. Under each
fop are logic, Y . . .
,,esses h e may be other intermediate events (intermediate J- . ~mt~ …
11edcace . be discovered as the mvesngauon goes on), basic iitrt’fl conunue to .
cs n1’Y that stop the chatn), undeveloped events (events that ,.-en al factors . .
. 0cs (caus .al bout which not enough 1s known to conunue the tree), ,,t enn or a
;0coosequ ( nts that are normally expected to occur). For example aJt _ 1 events eve ,
,0,1er<1"' din a hole and broke his leg, the broken leg is the top o 1 yee steppe ;rane01P0 ,
5
next tier of events includes the intermediate events “hole
‘fhe rree . · “G · c __ L ereoc. ” d “employee not paying attenuon. omg a step uu mer leads
[10,guarded an the answers to questions like “Why was the employee not
b ic ” ‘eots– d d,” Th · · co as • ;, Why was the hole not guar e . e mvestrgator must
. attenuon.
p,png f the accident scenarios in order to structure an accurate, fully
,oaly1e all 0
dereloped analytical tree. . ..
· a fault tree are logically directed through gates (see Exhibit 10.3). Alleveots 10
d” means that all outputs must occur. For example, if the top event lee “an gate
‘. ding an e-mail, then to make that top event occur, the computer must be
‘: 00 AND it must be connected to a phone line. Other things may have to
cum fth dfircil An”” th 1,,ppen as well, but both o ese must e te y occur. or gate means at
if anv one of the events on the second tier happens, the top event will happen.
Foe ~,ample, if the top event is making a million dollars, then to make it occur
)1lU could be a professional athlete OR win the lottery.
Once you determine the top event, the next step is to start tree construction.
The uee construction steps for an accident are:
I. Define the top event (accident, injury, or damage).
2 Investigate the accident. (Learn about the system, the management
structure, the accident, etc.)
3. Construct the tree. (Work from the top down asking why the top event
occurred.)
l. Develop causal factors. (The basic events-the boetom tier of the tree-
are causal factors.)
Pa11 Ill: A•alJliral T«bRiqm
The Analytic Tree Process
(Using MORT for Validation)
It is probably better to use MORT as a method of valid .
. . . . th . . •ling •noth
im·esnganon techruque an to use It as your pnmary techni ue er t}’Pe of
to ensure that you did not nuss an area that should have b! ·. It can help
and that the proper causal factors were determined If yo n 1nvestiga1’1J . . u consult th ‘ chart and find that some areas were nussed, the investigati e »!OR1 . . . . on can con . your invesnganon has been thorough, It will not take you Ion t tlnue. lf
the MORT chart. g o cornple,,
Example Scenario
Once again using the forklift-ladder accident, part of a tree that
al thi .d . ill d . could be used to an yze s aca. ent 1s . ustrate m Exhibit 10.7. The top event
is the warehouse supervtsor falling off the ladder. The next level is the
forklift hitting the ladder, and the next is three ways that a forklift could hit
a Ladder. The investigator asks questions about these three reasons to arri,..c
at the bonom tier-basic events or causal factors: “Why did the warehouse
supenisor not communicate what he was about to do to the supervisor of
the night shift? Was his failure to communicate a training issue, or did he
just decide to disregard the procedures? Why was the forklift traveling ,ith
an obstructed view?” If the accident investigator docs not know the answers
to these questions, he or she must interview witnesses, obtain documents, or
perform tests to find the answers.
\14
Chapter 1 O: Tree Analysis
. 7 E PROCESS FOR SCENARIO EXAMPLE
Part l/l: A nalytical Ttchniq11es
Summary
Many techniques are used in tree analysis . Each techni que works b
for some rypes of investigation than for others, and some tech . etter
. r f . . . mques ar
inappropnate ,or some rypes o mvest1gat1ons. The ma1·or b fi e ene u of us·
tree analysis is that trees are, for the most part, structured eas t mg . , Y o create and
easy to understand. You will be able to use one or more of thes h .’ e tee mques
to investigate almost every accident you encounter. (Please see th A . e ppenclix
for a sample Analyncal Tree Flowcharr.)
REVIEW QUESTIONS
1. Which type of gate requires that all outputs must occur?
2. What are the three categories of tree analysis? Briefly describe each.
3. What Is MORT?
4. What are the steps of tree construction?
s. Continue the analytic tree process for the example scenario (Exhibit 10.7),
116
U
. S . C H E M I C A L S A F E T Y A N D H A Z A R D I N V E S T I G A T I O N B O A R D
INVESTIGATION REPORT
REPORT NO. 2007-04-I-WV
SEPTEMBER 2008
LITTLE GENERAL STORE – PROPANE EXPLOSION
(Four Killed, Six Injured)
Photo courtesy of West Virginia State Fire Marshal
LITTLE GENERAL STORE, INC.
GHENT, WEST VIRGINIA
KEY ISSUES: JANUARY 30, 2007
• EMERGENCY EVACUATION
• HAZARDOUS MATERIALS INCIDENT TRAINING FOR FIREFIGHTERS
• 911 CALL CENTER RESOURCES
• PROPANE COMPANY PROCEDURES
• PROPANE SERVICE TECHNICIAN TRAINING
Little General Store September 2008
ii
Content
s
EXECUTIVE SUMMARY ………………………………………………………………………………………………………….. 1
KEY FINDINGS ………………………………………………………………………………………………………………………… 3
1.0 INTRODUCTION …………………………………………………………………………………………………………… 4
1.1 Summary ……………………………………………………………………………………………………………………….. 4
1.2 Investigative Process ……………………………………………………………………………………………………….. 7
1.3 Little General Store, Inc. ………………………………………………………………………………………………….. 8
1.4 Little General Store Propane Suppliers ………………………………………………………………………………. 8
1.5 West Virginia Emergency Service Organizations ………………………………………………………………… 9
1.6 Professional and Industry Organizations …………………………………………………………………………… 12
2.0 INCIDENT DESCRIPTION …………………………………………………………………………………………… 13
2.1 Events Preceding January 30, 2007 Explosion…………………………………………………………………… 13
2.2 Day of the Incident ………………………………………………………………………………………………………… 14
2.3 Response to the Propane Release …………………………………………………………………………………….. 17
3.0 PROPANE INCIDENT FREQUENCY ……………………………………………………………………………. 20
3.1 United States Hazardous Materials Incidents 2001 – 2006 ………………………………………………….. 20
3.2 Recent Propane Incidents ……………………………………………………………………………………………….. 21
4.0 PROPANE SYSTEM FUNDAMENTALS ………………………………………………………………………. 22
4.1 Propane Properties…………………………………………………………………………………………………………. 22
4.2 System Features…………………………………………………………………………………………………………….. 22
4.3 Propane Standards …………………………………………………………………………………………………………. 24
4.4 Propane Emergency Guidance ………………………………………………………………………………………… 25
5.0 INCIDENT AND EMERGENCY RESPONSE ANALYSIS ………………………………………………. 27
Little General Store September 2008
iii
5.1 Liquid Withdrawal Valve ……………………………………………………………………………………………….. 27
5.2 Propane Tank Placement ………………………………………………………………………………………………… 30
5.3 Propane Service Technician Training and Response ………………………………………………………….. 36
5.4 Fire Department Response ……………………………………………………………………………………………… 41
5.5 911 Emergency Call Center Response ……………………………………………………………………………… 44
6.0 REGULATORY ANALYSIS …………………………………………………………………………………………. 46
6.1 Occupational Safety and Health Administration ………………………………………………………………… 46
6.2 Environmental Protection Agency …………………………………………………………………………………… 46
6.3 West Virginia Fire Commission ………………………………………………………………………………………. 47
7.0 FINDINGS …………………………………………………………………………………………………………………… 51
8.0 CAUSES………………………………………………………………………………………………………………………. 53
9.0 RECOMMENDATIONS ……………………………………………………………………………………………….. 54
Governor and Legislature of the State of West Virginia …………………………………………………………………. 54
West Virginia Fire Commission ………………………………………………………………………………………………….. 54
West Virginia Office of Emergency Medical Services …………………………………………………………………… 54
National Fire Protection Association……………………………………………………………………………………………. 55
Association of Public-Safety Communications Officials ………………………………………………………………… 55
Propane Education and Research Council…………………………………………………………………………………….. 55
National Propane Gas Association ………………………………………………………………………………………………. 56
West Virginia E911 Council ………………………………………………………………………………………………………. 56
Ferrellgas …………………………………………………………………………………………………………………………………. 57
REFERENCES …………………………………………………………………………………………………………………………. 58
APPENDIX A FERRELLGAS INSTALLATION REVIEWS OF THE INCIDENT TANK……………… 60
APPENDIX B 911 CALL AND INITIAL FIRE DEPARTMENT DISPATCH ……………………………….. 63
Little General Store September 2008
iv
APPENDIX C RECENT PROPANE INCIDENTS ………………………………………………………………………. 66
RECENT PROPANE INCIDENTS …………………………………………………………………………………………….. 67
1.0 PROPANE RELEASE INCIDENTS ……………………………………………………………………………….. 67
1.1 Aberdeen, Washington …………………………………………………………………………………………………… 67
1.2 Lynchburg, Virginia ………………………………………………………………………………………………………. 68
1.3 Bristow, Virginia …………………………………………………………………………………………………………… 70
2.0 PROPANE RELEASE AND FIRE INCIDENTS ………………………………………………………………. 72
2.1 Sallis, Mississippi ………………………………………………………………………………………………………….. 72
2.2 Danville, Alabama …………………………………………………………………………………………………………. 73
APPENDIX D TABLE OF STATE REQUIREMENTS FOR PROPANE SERVICE TECHNICIANS . 75
APPENDIX E UNITED STATES EPA – 40 CFR 311 ………………………………………………………………….. 77
Little General Store September 2008
v
List of Figures
Figure 1. Aerial photograph of Little General store and surrounding plot.
…………………………………………. 4
Figure 2. Site plan of Little General Store. …………………………………………………………………………………….. 6
Figure 3. Liquid withdrawal valve. ……………………………………………………………………………………………… 15
Figure 4. Timeline of initial events ……………………………………………………………………………………………… 17
Figure 5. Timeline of incident response……………………………………………………………………………………….. 19
Figure 6. Typical ASME propane tank. ……………………………………………………………………………………….. 23
Figure 7. Incident valve plug (Telltale circled)………………………………………………………………………………. 27
Figure 8. Crack in valve seal……………………………………………………………………………………………………….. 29
Figure 9. Pre-incident photograph of the Ferrellgas tank. ………………………………………………………………. 30
List of Tables
Table 1. U.S. hazardous materials incidents 2001-2006 …………………………………………………………………. 20
Little General Store September 2008
vi
List of Acronyms and Abbreviations
ASME American Society of Mechanical Engineers
APCO Association of Public-Safety Communications Officials
CETP Certified Employee Training Program
CFR Code of Federal Regulations
CSB
U.S. Chemical Safety and Hazard Investigation Boar
d
DOT U.S. Department of Transportation
EMS Emergency Medical Services
EMT Emergency Medical Technicia
n
EOC Emergency Operations Center
EPA U.S. Environmental Protection Agency
FIR Ferrellgas Installation Review
HAZWOPER Hazardous Waste Operations and Emergency Response
HVAC Heating, Ventilation, and Air Conditioning
IC Incident Commander
LP Gas Liquefied Petroleum Gas
MSDS Material Safety Data Shee
t
NENA National Emergency Number Association
NFIRS National Fire Incident Reporting System
NFPA
National Fire Protection Association
NIMS National Incident Management System
NPGA National Propane Gas Association
OES Raleigh County Office of Emergency Services
OSHA U.S. Occupational Safety and Health Administration
PERC Propane Education and Research Council
RESA Regional Education Service Agency
SCGM Service Center General Manager
STARS Safety and Training Administrative Records System
WVC West Virginia Code
WVCSR West Virginia Code of State Rules
Little General Store September 2008
1
Executive Summary
On January 30, 2007, a propane explosion at the Little General Store in Ghent, West Virginia, killed tw
o
emergency responders and two propane service technicians, and injured six others. The explosion leveled
the store, destroyed a responding ambulance, and damaged other nearby vehicles.
On the day of the incident, a junior propane service technician employed by Appalachian Heating was
preparing to transfer liquid propane from an existing tank, owned by Ferrellgas, to a newly installed
replacement tank. The existing tank was installed in 1994 directly next to the store’s exterior back wall in
violation of West Virginia and U.S. Occupational Safety and Health Administration regulations.
When the technician removed a plug from the existing tank’s liquid withdrawal valve, liquid propane
unexpectedly released. For guidance, he called his supervisor, a lead technician, who was offsite
delivering propane. During this time propane continued releasing, forming a vapor cloud behind the
store. The tank’s placement next to the exterior wall and beneath the open roof overhang provided
a
direct path for the propane to enter the store.
About 15 minutes after the release began, the junior technician called 911. A captain from the Ghent
Volunteer Fire Department subsequently arrived and ordered the business to close. Little General
employees closed the store but remained inside. Additional emergency responders and the lead
technician also arrived at the scene. Witnesses reported seeing two responders and the two technicians in
the area of the tank, likely inside the propane vapor cloud, minutes before the explosion.
Minutes after the emergency responders and lead technician arrived, the propane inside the building
ignited. The resulting explosion killed the propane service technicians and two emergency responders
who were near the tank. The blast also injured four store employees inside the building as well as two
other emergency responders outside the store.
Little General Store September 2008
2
The CSB identified the following causes:
1. The Ferrellgas inspection and audit program did not identify the tank location as a hazard.
Consequently, the tank remained against the building for more than 10 years.
2. Appalachian Heating did not formally train the junior technician, and on the day of incident he was
working alone.
3. Emergency responders were not trained to recognize the need for immediate evacuation during liquid
propane releases.
The CSB makes recommendations to the governor and legislature of the State of West Virginia, the West
Virginia Fire Commission, the West Virginia Office of Emergency Medical Services, the National Fire
Protection Association, the Association of Public-Safety Communications Officials, the Propane
Education and Research Council, the National Propane Gas Association, the West Virginia E911 Council,
and Ferrellgas.
Little General Store September 2008
3
Key Findings
1. The propane service technicians, emergency responders, and store employees did not evacuate the
area as recommended by nationally accepted guidance for propane
emergencies.
2. A defect in the existing tank’s liquid withdrawal valve caused it to malfunction and remain in an open
position.
3. The junior propane service technician who was servicing the tank on the day of the incident had no
formal training and did not recognize the defect in the withdrawal valve. He was also working
unsupervised, even though he had been on the job for only one and a half months.
4. The placement of the 500-gallon propane tank against the building’s exterior back wall provided
releasing propane a direct path into the store’s interior.
5. The Occupational Safety and Health Administration’s and National Fire Protection Association’s
propane standards require training but do not include curricula, practical exercises, or knowledge
evaluation.
6. 911 operators in the United States lack propane emergency guidance to help them collect important
information from callers, offer life-saving advice, and convey relevant information to first responders.
7. Firefighters in West Virginia are required to attend a minimum of four hours of hazardous materials
emergency response training as part of their initial training sequence, but refresher training is not
required. The responding Ghent Volunteer Fire Department captain last attended a hazardous
materials response course in 1998.
8. Propane safety and emergency training is voluntary for fire department personnel in West Virginia.
None of the responders from the Ghent Volunteer Fire Department had specific propane emergency
training.
Little General Store September 2008
4
1.0 Introduction
1.1 Summary
At 10:53 am on January 30, 2007, a propane explosion leveled the Flat Top Little General Store (Little
General) in Ghent, Raleigh County, West Virginia (Figure 1). The explosion killed four and injured six.
The dead included two emergency responders (a fire department captain and an emergency medical
technician, both from the Ghent Volunteer Fire Department) and two Appalachian Heating propane
service technicians. The injured included the four Little General employees who remained inside the
store, and two other Ghent Volunteer Fire Department emergency responders.
Figure 1. Aerial photograph of Little General store and surrounding plot.
Little General Store September 2008
5
The morning of the explosion, a junior propane service technician1 (junior technician) from Appalachian
Heating was preparing to transfer liquid propane from an existing tank owned by Ferrellgas to a newly
installed tank2 owned by Thompson Gas and Electric Services (Thompson). The Ferrellgas propane tank
was installed in 1994 directly against the store’s exterior back wall (Figure 2). At about 10:25 am, the
junior technician, working alone, removed a threaded plug from the liquid withdrawal valve3 on the
Ferrellgas tank and liquid propane began flowing uncontrollably. Liquid propane sprayed upward,
against the roof overhang, and dense propane gas accumulated at ground level around the tank and the
foundation of the building. Over the next 25 minutes, the escaping propane entered the Little General
store through openings in the roof overhang.
Shortly after the release began, the junior technician called the lead technician to report the release and
seek guidance. At 10:40 am, the junior technician called 911 to report the emergency and summon help.
A captain and two emergency medical technicians from the Ghent Volunteer Fire Department were the
first to arrive, followed by the lead technician and two other emergency responders. Shortly after their
arrival, the propane in the store ignited, leveling it and killing two emergency responders (the fire captain
and one of the emergency medical technicians) and the two Appalachian Heating propane service
technicians.
1 The report discusses the activities of two propane service technicians: a junior technician, who had been
performing propane duties for one and a half months, and a lead technician, who had been performing propane
duties for one and a half years. “Junior” and “lead” are used in this report to differentiate the technicians’ relative
experience in propane service.
2 While commonly referred to as tanks, both of these were 500-gallon pressure vessels. The American Society of
Mechanical Engineers (ASME) publishes the Boiler and Pressure Vessel code; generally, stationary propane tanks
are considered unfired pressure vessels and manufactured in accordance with Section VIII of the code.
3 The liquid withdrawal valve was a RegO Chek-Lok valve model number 7572FC, which is no longer
manufactured. Although the RegO name is still used on propane equipment, the company that manufactured this
valve is no longer in business.
Little General Store September 2008
6
Fire departments from the neighboring communities of Beckley, Beaver, and Princeton responded to the
explosion. Later that day a team from the West Virginia Office of the State Fire Marshal arrived to
investigate, assisted by an agent from the U.S. Bureau of Alcohol, Tobacco, Firearms, and Explosives
(ATF).
Figure 2. Site plan of Little General Store.
Little General Store September 2008
7
1.2 Investigative Process
The CSB investigation team arrived at the incident scene on January 31. They joined the Incident
Command structure, in accordance with the National Incident Management System (NIMS),4 and began
on-scene investigation activities. On February 2, 2007, Incident Command demobilized after the State
Fire Marshal concluded that the incident was not a criminal act. The CSB investigation team remained,
and with the help of Little General management, protected and preserved evidence, moving it to a secure
storage locker.
The team interviewed employees of the companies involved, emergency responders, and officials from
the West Virginia Office of the State Fire Marshal; The West Virginia Division of Labor; the Raleigh
County Building Department; Regional Education Service Agency (RESA)5 Region I; the United States
Fire Academy; the ATF; the Beckley and Beaver, West Virginia Fire Departments; the Occupational
Safety and Health Administration (OSHA); the National Fire Protection Association (NFPA); the
National Propane Gas Association (NPGA); and the Propane Education and Research Council (PERC).
In addition, the CSB tested and examined the valve that released the propane from the Ferrellgas tank.
The test protocol included in situ examination and flow testing; removal of the liquid withdrawal valve
assembly (valve, tank nozzle, and dip tube); photography and examination of the valve and dip tube,
including removal of the dip tube; and dismantling and examination of the valve.
4 NIMS is a comprehensive approach to incident response management that provides a consistent, nationwide
template to enable all response entities to work in concert during incidents. Implementation of NIMS is required
by the US Department of Homeland Security in accordance with its authority in Homeland Security Presidential
Directive 5 “Management of Domestic Incidents.” The State Fire Marshal reestablished an incident command
system following the explosion.
5 Enacted by the West Virginia Legislature in 1972, RESA provides educational services to schools, including
technical, professional, operational, and programmatic services. In addition to school-based programs, RESA
coordinates much of West Virginia’s professional firefighter training program.
Little General Store September 2008
8
1.3 Little General Store, Inc.
Little General Store, Inc. operates 48 convenience stores throughout southern and central West Virginia
and western Virginia. The Ghent store was a combination gasoline station and convenience market. At
the time of the incident, the Ghent store and three others included pizzerias, which used propane for
cooking.
1.4 Little General Store Propane Suppliers
1.4.1 Southern Sun
Southern Sun supplied propane to Little General beginning in late 1994. Southern Sun was a family-
owned propane, heating oil, and ice supplier located in south central West Virginia. In 1996 Southern
Sun sold its propane operations to Ferrellgas, which became the supplier to Little General.
1.4.2
Ferrellgas
Ferrellgas, headquartered in Overland Park, Kansas, is the second-largest propane marketer in the United
States, with offices and customers in all 50 states. One of Ferrellgas’ business strategies is to “expand
operations through disciplined acquisitions and internal growth.” Since 1986 Ferrellgas has acquired 166
propane distributors throughout the United States.
Late in 2006, Little General initiated a change in propane suppliers from Ferrellgas to ThompsonGas
Propane Partners.
1.4.3 Thompson Gas and Electric Service, Inc.
Thompson is a privately held company that installs commercial and residential propane systems and
delivers propane in the eastern and southeastern United States.
Little General Store September 2008
9
1.4.4 Appalachian Heating
Appalachian Heating is a family-owned heating, ventilation, air conditioning (HVAC), and plumbing
company. Appalachian installs appliances, installs and maintains heating and cooling systems, and
installs and repairs plumbing. Appalachian entered into a contract with Thompson in August 2005 to
extend its business to include propane supply.
1.4.5 ThompsonGas Propane Partners, LLC
ThompsonGas Propane Partners is the limited liability company (LLC) formed between Thompson of
Hagerstown, Maryland, and Appalachian Heating of Bradley, West Virginia. Under the agreement
forming ThompsonGas Propane Partners, LLC, Appalachian Heating provides personnel to install
propane systems and deliver propane; Thompson provides equipment, bulk propane, and technical
support. The LLC has no employees.
1.5 West Virginia Emergency Service Organizations
1.5.1 West Virginia State Fire Commission
The West Virginia Fire Prevention and Control Act6 of 1975 established the West Virginia State Fire
Commission and granted the commission authority to promulgate and establish a state fire code.7 The
commission established the National Fire Codes8 as the minimum fire prevention and protection
requirements for the state.
6 West Virginia Code Chapter 29, Article 3 “The Fire Prevention and Control Act.”
7 West Virginia Legislative Rule Title 87, Series 1, “State Fire Code.”
8 NFPA publishes the National Fire Codes annually. The National Fire Codes are a collection of all NFPA’s
standards.
Little General Store September 2008
10
1.5.2 West Virginia State Fire Marshal
The West Virginia State Fire Marshal’s Office, overseen by the State Fire Commission, has four divisions
providing fire protection and regulatory services: regulation and licensing, public education, fire
investigation, and regional response. The Fire Marshal enforces the rules of the State Fire Commission
throughout West Virginia. Currently, the Fire Marshal’s Office employs 12 code inspectors and 11 fire
investigators.
1.5.3 Raleigh County Emergency Services
An executive group is responsible for emergency services management in Raleigh County, West Virginia.
Group membership includes county commissioners; the district attorney; mayors of municipalities; the
county sheriff; the Office of Emergency Services (OES); and the county public information officer.
Raleigh County emergency services encompass mutual aid management among municipal fire
departments (career and volunteer); police; emergency medical transport services; and community-based
support agencies. Services are coordinated under a unified command system at the Emergency
Operations Center (EOC) located in Beckley.
Local municipalities are responsible primarily for emergency response activities within their areas.
County resources are available from the OES and coordinated through the EOC when emergencies exceed
local response capabilities. The OES develops and maintains the Raleigh County Emergency Plan and
manages the county’s 911 emergency call center.
1.5.4 Ghent
Volunteer Fire Department
The Ghent Volunteer Fire Department was incorporated in 1973 and has 28 members providing fire
fighting, life protection, and ambulance service to residents in the Ghent area. Salaried Emergency
Little General Store September 2008
11
Medical Technicians (EMTs) are on duty at the fire station 24 hours a day, seven days a week.9 All
Ghent area emergency 911 calls are routed to the Raleigh County EOC where operators dispatch the
appropriate fire or emergency medical service.
1.5.5
West Virginia E911 Council
West Virginia established the E911 Council in 1986 to organize and implement the universal 911
emergency telephone number system. The council promotes, researches, plans, educates, develops
funding streams, and proposes state legislation to ensure reliable 911 call service operations.
Representatives from West Virginia’s 55 counties serve on the council, which meets monthly.
The council works closely with emergency response industry organizations including the National
Emergency Number Association (NENA) and the Association of Public-Safety Communicators Officials
(APCO). In addition, the council is involved in activities to promote and fund 911 systems statewide.
1.5.6
West Virginia Office of Emergency Medical Services
The West Virginia Office of Emergency Medical Services (EMS) is a division of the West Virginia State
Trauma and Emergency Care System, administered by the West Virginia Department of Health and
Human Resources. The Office of EMS was legislatively mandated in 1975 and tasked with operating a
comprehensive statewide EMS program. It oversees licensing for EMS agencies and training and
certification for responders.
9 The Ghent fire department comprises trained volunteer firefighters and paid emergency medical technicians.
Many of the emergency medical technicians, including those who responded to the Little General propane release,
also serve as volunteer firefighters.
Little General Store September 2008
12
1.6 Professional and Industry Organizations
1.6.1 National Fire Protection Association
The NFPA publishes consensus standards applicable to specific industries and activities, including the
propane industry and hazardous materials incident response. These standards, while voluntary unless
incorporated into state laws or regulations, provide safe practice guidelines for operations in the processes
they address. West Virginia adopted the National Fire Codes into the State Fire Code.10
1.6.2
Propane Education and Research Council
Congress created the Propane Education and Research Council (PERC) in the Propane Education and
Research Act of 1996. PERC’s purpose is to promote the safe use of propane energy in the United States.
Each gallon of odorized propane sold in the United States is assessed one-half of one cent to fund PERC.
PERC estimates revenue of $45,300,000 for the 2009 fiscal year.
PERC develops the curriculum for the Certified Employee Training Program (CETP), which is the
propane industry’s primary training method for service technicians, bulk plant operators, delivery truck
drivers, and customer service representatives.
1.6.3
National Propane Gas Association
Founded in 1931, the National Propane Gas Association (NPGA) is the trade association representing all
propane industry segments. It is located in Washington, D.C., and affiliated with 38 state and regional
associations, including the West Virginia Propane Gas Association.
The NPGA administers CETP throughout the United States.
10 The CSB determined that the National Fire Codes were adopted by the West Virginia State Fire Marshal prior to
1967, although the exact date is unknown.
Little General Store September 2008
13
2.0 Incident Description
2.1 Events Preceding January 30, 2007 Explosion
Significant events preceding the January 30, 2007, explosion at the Ghent Little General Store include
1. 1988 – Southern Sun buys the incident tank and places it in propane service.
2. Late 1994 – Southern Sun relocates the incident tank from its original installation to the Little General
Store in Ghent and installs it against the store’s exterior back wall.11
3. 1996 – Ferrellgas buys the Southern Sun propane business and performs a pre-acquisition review of
the business and operations, including an inspection of some of Southern Sun’s propane systems.
4. 1996 to 2007 – Ferrellgas drivers fill the incident tank about 14 times per year conducting about 140
pre-fill inspections.
5. April and June 2000 – Ferrellgas employees perform two Installation Reviews (Appendix A) on the
propane system at Little General.
6. August 2005 – Thompson Gas and Electric Service, Inc. and Appalachian Heating form the business
ThompsonGas Propane Partners, LLC.
11 At the time of the installation, West Virginia and OSHA required 500-gallon propane tanks to be at least 10 feet
from buildings; however, Raleigh County had no code enforcement mechanism at that time.
Little General Store September 2008
14
7. September 2006 – Appalachian Heating’s only propane service technician (the lead technician at the
scene of the incident) completes training and becomes fully qualified (in accordance with CETP) to
install propane tank systems.12
8. Fall 2006 – Little General Store, Inc., begins changing propane suppliers from Ferrellgas to
ThompsonGas Propane Partners.
9. December 2006 – Appalachian Heating reassigns an HVAC technician to the propane business (the
junior technician at the scene of the incident).
10. January 2007 – The two propane service technicians from Appalachian Heating began installing a
new Thompson Gas tank and associated equipment at Little General.
2.2 Day of the Incident
On the morning of Tuesday, January 30, the two propane service technicians traveled separately to the
Little General Store. At around 9:15 am, witnesses observed both of their trucks at the store. According
to Appalachian Heating, the day’s work plan was to transfer the propane from the existing Ferrellgas tank
to the newly installed Thompson tank and place the new propane system in service.
By 9:30 am the lead technician had left the store to make a delivery 31 miles northwest of Little General;
the junior technician remained alone at the store. For the next hour, the junior technician worked alone
while he prepared to transfer propane from the Ferrellgas tank to the Thompson tank.
12 Thompson trained the lead technician on sections 1, 2, and 4 of CETP. The lead technician also spent nearly a full
year working with experienced personnel from multiple Thompson offices.
Little General Store September 2008
15
Around 10:25 am, the junior technician removed a plug13 from the liquid withdrawal valve (Figure 3) on
the Ferrellgas tank. When he removed the plug, liquid propane unexpectedly began spraying from the
valve.
Figure 3. Liquid withdrawal valve.
At 10:28 am, the junior technician called14 the lead technician, who was still offsite, to report the problem.
In response, at 10:31 am, the lead technician called Thompson technical support for guidance. After the
call to technical support, the lead technician called the junior technician twice, most likely to relay the
guidance from technical support and to encourage him to call 911.
13 The valve plug has two functions: to prevent a propane leak if the valve seat leaks and to prevent foreign
materials’ entering the valve mechanism. The withdrawal valve itself is spring-loaded and self-closing and should
not leak; however, if it does, a telltale hole drilled through the threaded portion of the plug sprays a small amount
of propane giving an early warning of a malfunction prior to completely removing the plug.
14 All telephone calls discussed throughout this report were from cellular phones.
Plug
Valve Body
Valve sealing disk and washer
Little General Store September 2008
16
A deliveryman in the store at about 10:30 am told the CSB that he smelled a strong odor and that the store
employees’ eyes were watering. The Little General cashier, concerned about the odor, went outside to
check on the junior technician. The cashier told the CSB that she saw him standing between the tank and
his service truck within a dense vapor cloud,15 but that he told her he was “okay.”
At 10:40 am, the junior technician called 911 to summon help from emergency services (see transcript in
Appendix B). He reported to the 911 operator that he had a leaking propane tank at the Little General in
Ghent and needed the fire department’s assistance to secure the area. The 911 operator collected this
information, clarified the location, and ended the call.
After the 911 call, the junior technician called the lead technician again. The cashier checked on the
junior technician a second time, finding him in the same location between the tank and truck. Figure 4 is
an event timeline prior to the lead technician’s return.
15 Propane is a colorless gas. A liquid release from a tank forms a dense white cloud of visible propane liquid
droplets that then evaporate.
Little General Store September 2008
17
Figure 4. Timeline of initial events
2.3 Response to the Propane Release
After the junior technician’s call, Raleigh County Emergency Services dispatched the Ghent Volunteer
Fire Department to the Little General (Appendix B).
The first to arrive, at about 10:47 am,16 was a captain from the Ghent Volunteer Fire Department. The
captain assumed the role of incident commander (IC) in accordance with Ghent Volunteer Fire
Department guidelines.17
16 As no record of precisely what time the captain arrived at the scene exists, the CSB estimated it from other known
events.
17 Although functioning in accordance with the Ghent Volunteer Fire Department guidelines, the captain had no
formal incident command training and did not establish command in accordance with NIMS.
Little General Store September 2008
18
Shortly thereafter, two EMTs arrived in an ambulance. The surviving EMT told investigators that he saw
the IC at the store’s front door when he arrived. According to the Little General store manager, the IC
had ordered the business to close.18
The IC asked the EMTs to follow him behind the building to treat the junior technician for a burn on his
forearm. The EMTs examined the junior technician and determined that the burn was frostbite likely
from the releasing liquid propane.
Following the medical examination, the IC sent the EMTs to the front of the store to ensure that the
business had closed, gasoline was not being pumped, and no one was smoking. One EMT walked to the
store entrance and found it locked; however, he was able to speak to the Little General employees inside
to confirm that the business was closed. The Little General employees had hung a small sign on the door
to inform customers that the store was closed due to a gas leak. The EMT positioned himself in the
parking lot and proceeded to direct traffic and customers away from the property.
Just after 10:50 am, the lead technician returned to the store. He parked in front of the building and
walked to the area of the tanks.
A firefighter arrived in his personal vehicle to assist with the emergency and checked in with the IC. The
firefighter saw the IC and the two technicians working around the leaking tank; he could hear the
escaping propane. The IC ordered him to “make sure everybody’s out, okay?” However, as the
firefighter began walking to the front of the store, the propane ignited and exploded.
18 When ordered to close, the Little General employees turned off the gasoline pumps, locked the doors, and
remained in the building.
Little General Store September 2008
19
The explosion killed four near the tanks (the two technicians, the IC, and one EMT). The four workers
inside the building and the firefighter survived but sustained serious injuries. The EMT, who was
directing traffic and customers, suffered an arm injury.
The explosion leveled the building, destroyed a Ghent Volunteer Fire Department ambulance (cover
photograph), and damaged many parked vehicles. The force of the blast threw the Ferrellgas tank about
80 feet and the Thompson tank about 50 feet. Figure 5 is an event timeline of the initial emergency
response.
Figure 5. Timeline of incident response.
Little General Store September 2008
20
3.0 Propane Incident Frequency
3.1 United States Hazardous Materials Incidents 2001 – 2006
The National Fire Incident Reporting System (NFIRS) is the United States Fire Administration’s database
of fire and emergency incidents. The CSB reviewed the NFIRS database for the years 2001 to 2006 to
determine the frequency with which United States fire departments responded to propane incidents.
In this six-year period, the NFIRS database includes 36,744 hazardous materials responses in the United
States.19 Of these responses, 35.3 percent were responses to releases of gasoline, natural gas, or propane.
Incidents involving propane occur nearly once per day.
Table 1. U.S. hazardous materials incidents 2001-2006.
Gasoline
Natural
Gas Propane
Total
hazardous
materials
incidents
2001 658 468 157 4154
2002 817 627 210 4661
2003 1084 830 263 5904
2004 1149 863 234 6307
2005 1423 1151 301 8301
2006 1406 997 317 7417
Total 6537 4936 1482 36,744
19 As not all states require participation in NFIRS data collection, the total number of hazardous materials responses
may not be exhaustive; however, the CSB and the U.S. Fire Administration consider the types of responses
representative of the United States as a whole.
Little General Store September 2008
21
3.2
Recent Propane Incidents
Since the explosion in Ghent, several similar propane release incidents have occurred in the United States.
The CSB researched five incidents similar to the Little General incident (see Appendix C).
Three incidents involved only propane releases and two involved releases and flash fires. In the three
release-only incidents, both firefighters and propane service technicians responded. A propane service
technician secured one of the incident tanks, and firefighters trained in hazardous materials response
secured the other two. No injuries occurred in these incidents.
The two incidents involving releases and flash fires both occurred at sites where propane service
technicians were involved in tank-to-tank transfers of liquid propane. In these cases, fire departments
responded, and in both incidents destroyed the structures located nearest the propane tanks. One flash fire
injured a person onsite. The other killed one propane service technician and severely injured two others.
Firefighters, propane service technicians, or both regularly respond to propane release incidents. The
successful mitigation of a release depends on the training and experience of both groups.
Additionally, shortly before the Ghent incident, a liquid propane release and explosion occurred at the
Falk Corporation facility in Milwaukee, Wisconsin on December 6, 2006. Employees discovered a pool
of liquid propane near a propane pipeline. Most employees began evacuating; three remained to attempt
to repair the leak. About 10 minutes after the pool of propane had been discovered, it ignited, killing the
three employees and injuring over 40 others. OSHA cited Falk and its piping contractor for failing to
prepare employees for emergencies involving liquid propane.
Little General Store September 2008
22
4.0 Propane System Fundamentals
4.1 Propane Properties
Propane is used extensively as a fuel for cooking and heating in residential, retail, and commercial
applications.20 It is a gas at ambient temperature and pressure but is transported and stored as a liquid.
Propane-air mixtures can be explosive when the atmospheric concentration of propane is between 2.4 and
9.5 percent by volume. Because propane is odorless and colorless, small quantities of an odorizing
chemical (ethyl mercaptan) are added so leaks can be detected by smell.
Unlike natural gas, propane is heavier than air. Consequently, propane releases tend to accumulate at
ground level or in other low points such as pits and basements.
4.2 System Features
The propane system at Little General consisted of a stationary 500-gallon storage tank (Figure 6) and two
supply regulators to reduce the pressure of the propane for pizza ovens used in the store. The storage tank
had five connections:
1. A fill connection fitted with an internal check valve;
2. A liquid level gage to measure the quantity of propane in the tank;
3. A relief valve (safety device) to prevent over-pressurization of the tank;
4. A service valve to supply gaseous propane to the regulators; and
Little General Store September 2008
23
5. A liquid withdrawal valve fitted to an internal dip tube for removing liquid propane when necessary.
Figure 6. Typical ASME propane tank.
The liquid withdrawal valve incorporated several safety features:
1. A threaded valve plug to protect the valve from foreign material and provide a secondary seal if the
primary valve were to leak.
2. A telltale in the threaded area on the valve plug. The telltale is exposed before the plug is completely
removed. The telltale releases propane if the primary valve leaks allowing a propane service
technician to check for valve leaks before completely removing the plug.
3. A special connector that, when screwed on, opens the valve.
20 According to PERC, about 17.5 million U.S. households use propane as a heating or cooking fuel, excluding
grills. About 3,500 companies market propane around the country.
Little General Store September 2008
24
4. An excess flow disk in the valve to limit flow should flow exceed a preset limit, such as during a hose
rupture.
4.3 Propane Standards
4.3.1 Occupational Safety and Health Administration
The OSHA standard Storage and Handling of Liquefied Petroleum Gases21 establishes specific
requirements for most workplaces22 storing and handling liquefied petroleum gases (LP gases)23,
including propane.
The standard’s requirements include, but are not limited to, design, operation, technician training, safety
devices, and tank placement. A 500-gallon tank, like the one at Little General, must be located at least 10
feet from buildings.
4.3.2 NFPA Propane Standards
Two NFPA standards address propane: the National Fuel Gas Code (NFPA 54), which applies to fixed
gas-consuming (including propane) equipment attached to piping, and the Liquefied Petroleum Gas Code
(NFPA 58), which applies to the storage, handling, transportation, and use of LP gas. NFPA 58 Chapter 6
addresses installing LP gas systems, including requirements for tank locations. NFPA 58 also requires a
500-gallon tank to be at least 10 feet from commercial and residential buildings.
21 29 CFR 1910.110.
22 Little General is covered by OSHA’s occupational safety and health standards.
23 Liquefied petroleum gas is the general term for mixtures of hydrocarbon gases liquefied by pressure and used for
cooking, heating, and vehicle fuel. Liquefied petroleum gas can be propane, butane, propylene, butylene or any
mixture of these compounds.
Little General Store September 2008
25
4.4 Propane Emergency Guidance
4.4.1 Guidance for Consumers
Propane companies (including Ferrellgas and Thompson), NPGA, and PERC distribute literature and
publish information on their websites guiding consumers on actions to take during propane emergencies
such as leaks and releases. If consumers detect a propane leak, this guidance recommends that they:
• Extinguish smoking materials and discontinue the use of lights, appliances, phones, or any electrical or
spark producing device.
• Evacuate the area or building immediately.
• If possible, turn off the gas supply at the tank or cylinder.
• Go to a neighbor’s house and call the propane company or 911.
• Stay out of the area or building where the leak was detected.
4.4.2 Guidance for Emergency Responders
Propane Emergencies, published for the NPGA and PERC, is the primary text for training emergency
responders to handle
propane emergencies.
The textbook outlines a process wherein the IC should
perform a series of actions prior to attempting remediation of the propane emergency. The first step the
textbook specifies is site control; it states that the IC must keep all responders and members of the public
clear of the hazard until the scene and the hazard are fully understood.
The US DOT Emergency Response Guidebook recommends an immediate evacuation to at least 330 feet
in all directions and ½-mile downwind for large spills. Responders are cautioned to keep all members of
Little General Store September 2008
26
the public away from the area surrounding the point of the release. In addition, the Guidebook states that
propane gas is heavier than air and may settle in low or confined areas.
Little General Store September 2008
27
5.0 Incident and Emergency Response Analysis
5.1 Liquid Withdrawal Valve
Liquid withdrawal valves are rarely used. Technicians only open them when tanks must be completely
emptied of liquid. The liquid withdrawal valves on most propane tanks may never be used. At Little
General, the day of the incident was either the first or the second time that the plug had been removed
from the liquid withdrawal valve in the tank’s 19 years of propane service.
When the junior technician removed the plug from the liquid withdrawal valve, liquid propane released
uncontrollably. The plug, though, has a telltale drilled through the threaded portion (Figure 7) which
should have released a small stream of propane once the plug was partially backed out. This should have
alerted the technician that the valve was leaking.
Figure 7. Incident valve plug (Telltale circled).
Little General Store September 2008
28
The CSB determined that the junior technician likely was unaware of the existence and function of the
telltale due to inexperience and lack of training and removed the plug completely. Although the CSB
determined that it is highly unlikely the telltale hole was obstructed at the time of the incident, CETP
training could have made the junior technician aware of and prepared for the possibility of an obstruction
of the telltale. The CETP section on liquid withdrawal valves states:
In some cases, a damaged seat may allow an excessive amount of liquid to be discharged when
the closing cap [plug]24 is loosened. A bleed hole [telltale] in the closing cap has been provided
to vent the liquid before the cap is completely unscrewed. If a significant amount of liquid
continues to be blown from under the closing cap for more than 30 seconds, it can be assumed
that the internal seat will not prevent a dangerous amount of gas from escaping. IF IN DOUBT,
DO NOT REMOVE THE CLOSING CAP. [CETP 2.2.13]
In addition to the information on the valve above, the CETP procedure for transferring liquid propane
from a tank to a truck states:25
While removing the plug, if an excessive volume of propane leaks from the valve, reinstall the
plug and follow manufacturer’s instructions. [CETP 2.2.13]
The CSB initially tested and examined the tank and valve in its “as-found” condition. The CSB
determined that the liquid withdrawal valve leaked when the tank was pressurized.26 Following this
examination and test, the CSB removed the valve from the tank, performed additional examinations and
24 Plug (external threads) closures are no longer used with liquid withdrawal valves; therefore CETP refers only to
cap (internal threads) closures.
25 CETP contains step-by-step procedures for liquid propane transfer from a stationary tank to a tank truck. Liquid
propane transfer from a stationary tank to a second stationary tank is not described in CETP.
26 Nitrogen was used to pressurize the tank since it is inert and allowed this test to be conducted safely.
Little General Store September 2008
29
tests, and partially disassembled the valve. The CSB found the valve jammed, likely because of a
manufacturing defect,27 and the valve seal cracked (Figure 8). With the valve jammed in an open
position, only the plug stopped the liquid propane from escaping. When the junior technician removed
the plug, propane began uncontrollably releasing. Cracks in the seal material likely prevented the excess
flow feature from limiting the release.
Figure 8. Crack in valve seal.
27 A hole bored through the center of the lower guide through which the valve stem moves was determined to be too
small for the valve stem to move freely.
Little General Store September 2008
30
5.2 Propane Tank Placement
In 1994, Southern Sun installed the 500-gallon propane tank directly against the Little General store’s
exterior back wall. The tank remained in this position and in operation until the day of the incident. The
position of the tank was contrary to both the West Virginia Fire Code and the OSHA LP gas standard.
Figure 9 shows a 2003 photograph of the building with the propane tank visible behind a wooden fence.
Photo courtesy of Little General Store, Inc.
Figure 9. Pre-incident photograph of the Ferrellgas tank.
Multiple witnesses told the CSB that Southern Sun placed the tank against the back wall. Paint patterns
on the tank corroborate this: the outward facing side of the tank was painted with the Ferrellgas color and
logo, while the side of the tank facing the building was painted with the Southern Sun name and logo
(Figure 10). This suggests that Ferrellgas painted only the side of the tank that was accessible.
Little General Store September 2008
31
Figure 10. Incident tank side facing building.
During the release, witnesses described seeing billowing white clouds striking the building and roof
overhang and cascading toward the ground, which is consistent with the known behavior of propane
vapor following a liquid release. The peaked roof of the store overhung the exterior walls. The overhang
was open to allow ventilation in the store’s attic space; restroom exhaust ducts terminated inside the
overhang directly above the incident tank. The open overhang and restroom exhaust ducts provided a
direct pathway for propane to enter the building. The CSB found that if the tank were placed at least 10
feet from the store, propane likely would not have entered the store in large quantities.
5.2.1 Southern Sun Acquisition
Ferrellgas’ acquisition of Southern Sun included a pre-acquisition review of Southern Sun’s business and
operations, which included inspections of propane systems at select Southern Sun customer locations. In
addition, Ferrellgas was allowed to identify and request correction of any deficiencies discovered in the
year following the acquisition. According to Southern Sun, only two propane tanks were returned, both
Little General Store September 2008
32
because of illegible or missing tank nameplates. Ferrellgas first filled the tank at the Little General soon
after the acquisition, and did not identify the tank placement as deficient.28
5.2.2 Ferrellgas Inspection and Quality Review Program
5.2.2.1 Ferrellgas Inspections
The Ferrellgas inspection program for propane systems at customer locations consists of two separate
activities: the driver’s pre-fill inspection, where delivery personnel visually inspect the system and the
Ferrellgas Installation Review (FIR), where an experienced service technician visually inspects the
system and records the findings.
1) Drivers’ Pre-Fill Inspection
Drivers perform pre-fill inspections prior to each propane delivery at a customer location. This visual
inspection includes the tank, visible piping, regulators, and the area around the tank. Ferrellgas trains
drivers to inspect the installation for potential deficiencies such as tank damage or corrosion,
accumulations of combustible materials near the tank, fitness of piping and regulators, tank placement and
new structures near the tank, and tank labeling. The driver does not record the results of the pre-fill
inspection, but does have the authority to remove the system from service upon finding unsafe conditions.
The driver must also report unsafe conditions to management so that a service technician can fix any
noted problems. Ferrellgas training materials for delivery drivers discuss tank placement and clearly state
that 500-gallon propane tanks must be at least 10 feet from buildings.
28 The week prior to the incident, Appalachian Heating had also replaced another Ferrellgas tank that was placed less
than 10 feet from a different Little General Store location.
Little General Store September 2008
33
Ferrellgas filled the tank at Little General Store about 14 times per year; in the 10 years that Ferrellgas
owned the tank, drivers would have performed more than 100 pre-fill inspections. The CSB interviewed
several drivers who filled the Little General tank and performed the required inspections. All drivers
stated that they believed the tank was permitted to be directly against the building’s exterior back wall
either because it was “grandfathered,”29 the installer had received a variance from local authorities,30 or
there was no other place to install the tank.
2) Ferrellgas Installation Review
The Ferrellgas Installation Review (FIR)31 is a visual inspection of the installation at a customer’s site.
The FIR includes the same topics as the driver’s pre-fill inspection; however, the inspector records the
results either on a form or handheld computer. The service center receives the results of the FIR, records
deficiencies, and dispatches service technicians to repair noted problems.
Ferrellgas performed two FIRs on the Little General Store installation in 2000, four years after acquiring
the tank (Appendix A). These inspections were two months apart; neither inspector noted the placement
of the tank on the form.32 Both inspectors told CSB that they could remember neither the inspections nor
the reasons for not reporting the tank’s placement in the FIR.
29 They believed that the tank installation predated the 10-foot clearance requirement, and was permitted to remain
next to the building under some previous set of rules.
30 Local authorities issue formal waivers from code requirements such as the tank placement distances, which are
known as variances. Ferrellgas training materials reference the possibility of obtaining a variance from local
authorities; however, a variance was not issued for the placement of the tank at Little General.
31 Ferrellgas called this inspection program the FIR during the time it owned the incident tank. The review has been
renamed and is now called the Product Installation Review. The methodology for inspection has not changed.
32 Checking the location of an installed tank is step one in the FIR procedures, which states that tanks that do not
meet the 10-foot placement requirement must have a written variance from local authorities.
Little General Store September 2008
34
5.2.2.2 Ferrellgas Quality Review Program
Ferrellgas’s Quality Review program33 comprises two annual internal audits; each reviews safety and
quality at a Ferrellgas service center location. Each service center’s general manager (SCGM) conducts
one audit (the SCGM audit), and a regional technical analyst conducts the other (the regional technical
analyst audit).
1. Service Center General Manager’s Audit
Since 2005, Ferrellgas has required its SCGM to perform quarterly audits of his service center. Each
SCGM uses a standard checklist-based procedure to audit specific segments of the business according to
the following schedule:
• First Quarter: Administration, including delivery, installation, and training documentation, operating
procedures, and inspection results;
• Second Quarter: Plant operations, including piping, tanks, system and container integrity, and operator
proficiency;
• Third Quarter: Delivery operations, including delivery trucks’ integrity and roadworthiness, and driver
proficiency; and
• Fourth Quarter: Customer installations, including inspection of 10 old and 10 new systems.34
33 The Quality Review program is independent of the FIR and the driver’s pre-fill inspection. Neither audit is
scheduled to coincide with a system installation or propane delivery.
34 New installations might be placed by Ferrellgas during the year or be acquired by buying other preexisting
propane systems.
Little General Store September 2008
35
The CSB requested all SCGM audits from the Princeton and Beckley, West Virginia, service centers for
the 10 years preceding the incident. Ferrellgas provided SCGM audits from 2005 and 2006 but provided
no documentation of any customer installation audits.
2. Regional Technical Analyst Audit
At least once annually, regional technical analysts (known as operations managers prior to 2005) conduct
compliance and safety audits at Ferrellgas service centers in their respective regions. The audit is similar
to the SCGM quarterly audit in that the technical analyst examines delivery operations, administration,
plant operations, and customer installations, using a checklist to guide the audit and record the results.
The only apparent difference is that the technical analyst performs all four segments of the audit within
one to two days and the SCGM examines one segment per quarter.
5.2.3 Ferrellgas Inspection and Quality Review Program Analysis
At least six Ferrellgas delivery drivers and service technicians were in a position to identify and report the
incorrect placement of the tank behind Little General; none did. Although Ferrellgas training materials
discuss correct tank placement and list tank placement as an inspection criterion for which drivers and
technicians may remove tanks from service, the employees who filled and inspected the tank believed that
its location was approved.35
The SCGM customer installation audits for the Beckley and Princeton offices were likely not conducted.36
Had managers performed these audits, it is still possible that because of the sampling methods used they
35 All Ferrellgas drivers and technicians reported to the CSB that they had received training in accordance with
Ferrellgas procedures.
36 Ferrellgas was unable to provide any documentation of customer installation audits for the ten years preceding the
incident.
Little General Store September 2008
36
would not have detected the improper installation at Little General. Ferrellgas’ instructions on
conducting audits merely recommend SCGMs inspect customer installations that were filled on the day
prior to the review or inspect those that the manager is driving past. These methods of sampling provide
no way to systematically review all installations over time.
SCGMs’ and regional technical analysts’ audits focus primarily on identifying and correcting deficiencies
at facilities and customer locations. Only those installations that are spot-checked during the quality
review are corrected; therefore broader systemic problems may remain undetected. In contrast, generally
recognized and accepted practices for auditing safety management systems focus on assessing and
verifying system effectiveness.37 These practices include using statistical sampling, trend analysis,
management system review, and corrective action to detect systemic problems and ensure ongoing hazard
control throughout the organization.38
5.3 Propane Service Technician Training and Response
5.3.1 Propane Service Technician Training
5.3.1.1 Federal and State Training Requirements
The United States Department of Transportation (DOT) and OSHA regulations have limited propane
service technician training requirements. DOT requires training only for propane personnel who engage
in transportation activities. OSHA requires training for propane service technicians but does not elaborate
on this requirement (Section 6.1).
37 Standard safety texts describe audits as systematic, independent reviews that determine conformance to company
and industry standards whereas inspections are physical inspection of installations and facilities.
38 CCPS, 1993; Petersen, D., 1996; Weinstein, M.B., 1997
Little General Store September 2008
37
Some form of state-mandated training and/or testing requirement for propane service technicians exists in
14 states, and 10 of those have specific training requirements; however, the depth of training required
varies broadly. No states require emergency response training for propane service technicians.
Three states require licensing for propane service technicians. States use licensing to establish and
enforce minimum competency standards for technicians; collect funds for inspection, compliance, and
training programs; and communicate technician qualifications to consumers. In the states that license
propane service technicians, CETP is an accepted training program. Additionally, at least 10 states have
licensing programs applicable to propane and natural gas appliance installers. Recurrent licensing
requirements, based on state-established standards of training, testing, and/or experience, aim to ensure
that only qualified personnel work on indoor propane and natural gas systems.
West Virginia does not require propane service technicians to be trained, certified, or licensed. Appendix
D contains a table of the state requirements.39
5.3.1.2 NFPA 58 Training Recommendations
NFPA 58 recommends training for propane industry employees who perform activities within its scope,
which includes those who transport and transfer LP gas. The standard states that these employees “shall
be trained in proper handling procedures. Refresher training must occur at least every 3 years. The
training shall be documented.”40 An appendix note states that refresher training may be less intensive
than original training. However, NFPA 58 does not elaborate on what constitutes initial training. It
includes no guidance suggesting a scope, timeframe, or testing component for facilitating an enforceable
training standard for states that have adopted NFPA 58, including West Virginia.
39 Information on state training requirements was provided to the CSB by PERC.
40 NFPA 58 – 2008 Section 4.4 “Qualifications of Personnel.”
Little General Store September 2008
38
5.3.1.3 Propane Industry Training Program
PERC and the NPGA developed its comprehensive training curriculum for propane industry employees,
CETP, in 1988. Although CETP is voluntary in most states, many employers in the propane industry use
it to train and certify employees. PERC offers the task-based modular curriculum both in instructor-led
settings and as a computerized self-teaching program, and sells course materials to employers for in-
house training and preparation for CETP certification tests.
CETP consists of eight modules. Applicable training modules include
• Propane “basics,” including properties of propane, industry standards and organizations, bulk
plant and delivery vehicle identification, safe work practices, and reacting to interruptions of
customer gas service;
• Propane delivery, including equipment and operating procedures, loading and unloading,
inspecting delivery trucks and installed propane systems, responding to customer calls, and
evacuating propane tanks; and
• Propane vapor distribution system installation, including transporting tanks and cylinders,
installing aboveground and underground tanks, installing regulators and piping, performing leak
checks, documenting installations, and communicating safety information to customers.
Employers may select topics as appropriate to address job tasks for individual employees, who may then
take certification exams in those areas for which they have completed training. For example, customer
service representatives often complete the propane basics module, and delivery drivers complete the
Little General Store September 2008
39
propane basics and propane delivery modules. Companies may also augment the curriculum to include
company-specific policies and information.41
An important part of the CETP curriculum is its skills testing component. The training program itemizes
basic propane procedures; within 12 months of passing the CETP certification exam, certification
candidates must demonstrate these procedures and be evaluated. The CETP curriculum includes skills
assessment forms; a trainer or other supervisor with sufficient experience must observe a certification
candidate perform or simulate procedures correctly, document performance evaluations for each set of
procedures, and forward the documentation to PERC before the candidate is certified to work alone.
The junior technician at Little General was preparing for a tank-to-tank transfer of liquid propane.
Although a CETP module provides step-by-step procedures for liquid transfers to and from bulk delivery
trucks, such procedures do not exist for tank-to-tank transfers. While some propane companies may
develop their own instructions for this task, the junior technician had no written procedure for his
activities on the day of the incident.
5.3.1.4 Propane Service Technician Emergency Response Training
A propane incident involving fire department response occurs nearly once per day in the United States
(see Table 1). The CSB found that propane service technicians routinely respond to these emergencies.
When the lead technician returned to the Little General store in response to the emergency, he was
functioning as a specialist responder, with expertise in propane and propane systems. By the time he
arrived, a hazardous materials release had been underway for at least 20 minutes. However, his training
had not included an emergency response component.
41 PERC offers the Safety and Training Administrative Records System (STARS), a computer-based program for
record keeping and training management, which also allows employers to add their own policies to modules.
Little General Store September 2008
40
The CSB found a number of other incidents in which propane service technicians assisted first responders
in emergencies involving their companies’ equipment and product. First responders consider propane
service technicians to be “product and container specialists” who have more experience dealing with
propane and its systems than firefighters, police, paramedics, and EMTs.42 Although propane service
technicians are not trained to take action to mitigate a hazardous materials emergency without specific
training,43 emergency responders often seek their advice and assistance.
PERC and the NPGA provide a Propane Emergencies training curriculum for first responders that
addresses the role of propane service technicians in the incident command structure (see Section 5.3).
However, CETP only briefly addresses emergencies at bulk plants and traffic accidents involving delivery
vehicles. No CETP module specifically addresses the potential emergencies service technicians
encounter or the basics of hazardous materials incident response.
5.3.2 Thompson Gas/Appalachian Heating Employee Training
Thompson Gas Propane Partners, LLC required all propane service technicians to be CETP-trained to the
appropriate level for their jobs. The 2005 contract with Appalachian Heating defines responsibilities,
including employee training: Appalachian is responsible for ensuring that all employees working with
propane are qualified “as defined in NFPA 54 and 58” and CETP-trained for their particular jobs. The
specific requirements include completion of CETP 1.0, Basic Principles and Practices, within 30 calendar
days of first working for Appalachian/Thompson, and completion of other applicable modules within one
year. Additionally, Appalachian Heating must maintain all training records and certifications, provide
ongoing training, and update the staff list for Thompson Gas before any new employees begin working.
42 Propane Emergencies, PERC and NPGA, 3rd edition, page 44.
43 HAZWOPER Standard, 29 CFR 1910.120(q).
Little General Store September 2008
41
Thompson provided CETP training for Appalachian’s lead technician in 2006, including the basic
principles, delivery, and vapor distribution system installation modules. Experienced Thompson trainers
directly supervised his performance of propane service and delivery tasks for nearly a year. This was
combined with the CETP classroom training to ensure that he was able to demonstrate proficiency and
pass certification examinations prior to working unsupervised.
The junior technician began working in the propane business in mid-December, about 45 days prior to the
incident. However, Appalachian did not inform Thompson of the new employee. At the time of the
incident the junior technician had not trained with CETP and had not received any other formal propane
service training.
Had Appalachian Heating implemented the training procedures required by the contract, the junior
technician would have, via CETP, become familiar with a range of procedures and situations that might
occur in the course of working with propane including operation of the liquid withdrawal valve.
5.4 Fire Department Response
The IC arrived at the Little General store shortly after the initial dispatch call. In the approximately five
minutes from his arrival until the explosion at 10:53 am, the IC took several actions. He:
• Assessed the frostbite injury to the junior technician;
• Ordered the business to close;
• Directed the EMTs to the rear of the building to treat the junior technician’s frostbite;
• Ordered the EMTs to ensure that the business was closed, that no one was smoking, and that no
gasoline was being pumped; and
• Ordered the firefighter to ensure that everyone was out of the building.
Little General Store September 2008
42
Guidance for emergency responders in hazardous materials emergencies recommends evacuating and
evaluating the situation from a safe distance as the first task. However, the IC’s final direction, to ensure
that everyone was out of the building, came too late. Within about 30 seconds of the order, the propane
ignited and the building exploded.
Based on witness statements and the IC’s known actions, the CSB determined that the IC allowed the
propane service technician to try to correct the propane release while the IC managed the scene.
However, the IC was likely unaware of two critical facts:
• The junior technician, upon whom the IC was relying to correct the release, was neither trained
nor experienced in propane transfer operations and propane emergencies (Section 5.3).
• The building was filling with a flammable mixture of propane and air. While the IC was
concerned with ignition sources outside such as cigarettes and vehicles, ignition sources inside
the building remained uncontrolled during the release.
Without knowledge of these facts, the IC was likely unable to fully understand the severity of the
emergency at Little General, even though he was fully trained in accordance with the rules of the West
Virginia Fire Commission.
5.4.1 Hazardous Materials Incident Training for Firefighters
West Virginia requires all firefighters to receive basic training in hazardous materials incident response.
This training is required prior to working as a firefighter and therefore is generally conducted only once,
Little General Store September 2008
43
early in a firefighter’s career. In contrast, chemical facilities and refineries in the United States that
maintain hazardous materials response teams are required to train team members annually.44
The required hazardous materials incident response training teaches firefighters to identify hazardous
materials and incidents involving hazardous materials, use DOT’s Emergency Response Guidebook, and
recognize when additional resources are necessary. The training instructs firefighters to take only
defensive actions when encountering a hazardous materials incident; it does not give firefighters the
knowledge or skills to enter the area of the release and attempt to stop the release. This action is reserved
for organized hazardous materials response teams.
The IC at Little General attended a hazardous materials incident response course in 1998. This was a
higher-level course than required (see Section 6.3 for requirements), and included topics such as incident
risk assessment, selection of personal protective equipment, and control techniques, and a practical
exercise. However, this training occurred nine years before the Little General incident, and the CSB
found no documentation that the IC had attended refresher training or practical exercises. Unrehearsed
knowledge decays over time; hazardous materials incident response skills acquired in training are more
difficult to retain without regular refresher training or practice.45
5.4.2 Propane Industry Training for Emergency Responders
The Propane Emergencies training program offered by PERC and the NPGA is free to fire departments
around the United States. The program trains first responders to recognize the specific hazards of a
44 29 CFR 1910.120 requires industrial hazardous materials incident responders to attend eight hours of refresher
training annually.
45 For a full discussion of skill loss, see Arthur Jr., W., Bennett Jr., W., Stanush, P., & McNelly, T. (1998). “Factors
That Influence Skill Decay and Retention: A Quantitative Review and Analysis,” Human Performance, Vol. 11,
pp. 57-101.
Little General Store September 2008
44
propane release and treat it as a hazardous materials incident. It emphasizes the importance of evacuating
the incident area and evaluating the hazards from a safe distance prior to taking other actions.
Additionally, it addresses incident command structure implementation and the advisory role of propane
service technicians. None of the Ghent Volunteer Firefighters had participated in this program.
5.5 911 Emergency Call Center Response
At many 911 call centers around the United States, operators use quick-reference guide cards to help them
evaluate caller emergencies, gather pertinent incident details, and convey life-saving information to
callers. Several organizations in the United States produce pre-written guide cards for 911 centers to
purchase.46 These guide cards provide easy-to-access information for operators, including questions
relevant to the emergency and instructions to be given to the caller prior to emergency services’ arrival.
Specific sets of guide cards exist for health and injury, fire service, and police response emergencies.
Raleigh County 911 uses guide cards for health emergencies and injuries only.
At 10:40 am on the day of the incident, the propane service technician called Raleigh County 911 to
report the release and summon emergency services. The operator who received the call did not have a
guide card or protocol to help evaluate the situation, collect pertinent information, and provide guidance
to the caller.
46 Guide cards are available as printed or electronic references and kept at each operator’s station.
Little General Store September 2008
45
The propane industry developed a model questionnaire and script to use in situations where customers
report propane emergencies such as leaks or releases. The questionnaire prompts personnel answering
calls to ask questions such as
• Where is the leak?
• Do you hear gas escaping?
• Is the leak near any building?
• Is there an odor of gas in the building?
An affirmative answer to these and other questions prompts the gas company operator to read a script that
instructs the caller to eliminate ignition sources, evacuate the building to a safe distance, and wait for gas
professionals or fire service personnel.
Equipping 911 operators with such a prewritten guide can potentially improve safety by initiating
important first response actions such as evacuation.
Little General Store September 2008
46
6.0 Regulatory Analysis
6.1 Occupational Safety and Health Administration
The OSHA standard regulating propane systems requires that all personnel “performing installation,
removal, operation, and maintenance of propane equipment shall be properly trained in such function.”47
However, neither the standard, its preamble, nor letters of interpretation define or interpret “properly
trained,” nor does it provide for specific training standards, requirements for recurrent training, written
testing, or skills testing. OSHA only requires employers to develop training programs they deem
appropriate without providing any performance criteria such as those under the industry’s program.
Additionally, the OSHA standard requires that owners of propane tank(s) be notified prior to work being
done on their tank(s) by others.48 Although Appalachian Heating employees removed the liquid
withdrawal valve plug from the Ferrellgas tank, Appalachian Heating did not notify Ferrellgas.
6.2 Environmental Protection Agency
Firefighters and emergency medical personnel are typically the first responders to hazardous materials
spills and releases, like the incident at Little General. However, in 26 states (including West Virginia),
state and municipal employees are not covered by the OSHA health and safety regulations that require
first responder hazardous materials training.49 In an attempt to close the gap, the U.S. Environmental
Protection Agency (EPA) promulgated the Worker Protection standard (Appendix E)50 for municipal and
state emergency responders.
47 29 CFR 1910.110(b)(16).
48 29 CFR 1910.110(b)(14)(ii).
49 See CSB Report No. 2006-03-I-FL “Bethune Point Wastewater Treatment Plant” for a full discussion.
5040 CFR 311.
Little General Store September 2008
47
The Worker Protection standard requires states and municipalities to prepare employees for hazardous
materials emergencies in accordance with OSHA’s Hazardous Waste Operations and Emergency
Response standard.51 The standard’s requirements apply to career and volunteer fire companies, and
public emergency medical response agencies. The United States Fire Administration and PERC52 both
cite the Worker Protection standard in their incident response training courses for firefighters. The rules
for firefighters and emergency medical personnel in West Virginia, however, do not specifically
incorporate the EPA requirements under the Worker Protection standard.
6.3
West Virginia Fire Commission
The primary role of the Fire Commission in West Virginia is the promulgation of the state fire code and
certification requirements for fire departments.
6.3.1 Requirements for Fire Departments53
The West Virginia Fire Commission certifies fire departments every five years based on
• location, size and boundaries of the fire protection district;
• number of active and available personnel and their level of training;
• quantity and type of equipment;
• administrative procedures and policies including chain-of-command;
51 29 CFR 1910.120 – Hazardous Waste Operations and Emergency Response generally requires employers to
develop emergency response and health and safety plans that address worker training, preparedness, and health
monitoring. The regulation applies to both long-term hazardous waste clean-up operations in addition to
emergency response for incidents involving any hazardous material (not limited to hazardous wastes). The
standard requires initial training and annual refresher training.
52 Propane Emergencies, 3rd ed., p. 41.
53WVC 29-3-2.
Little General Store September 2008
48
• 911 service and mutual aid agreements for the area served;
• the department’s record of response to emergency calls; and
• recordkeeping and NFIRS reporting.54
On April 1, 2005, the Fire Commission published an administrative policy to assist it in its evaluation of
fire departments.55 The policy addresses certifying and training active personnel in fire departments,
requiring them to attend Firefighter Level I, first aid, cardio-pulmonary resuscitation (CPR), and
hazardous materials response training.
The Fire Commission’s policy specifically requires the following for hazardous materials response:
Fire department members and officers shall be trained in NFPA Standard No. 472, Professional
Competence of Responders to Hazardous Materials Incidents.56 All Fire Department personnel
shall be certified at the Awareness level. Fire Department personnel that operate at hazardous
materials incidents must minimally meet U.S. Environmental Protection Agency and U.S.
Occupational Safety and Health Administration requirements for response at the Operations level.
Additional training should be at the discretion of the fire chief based on possible involvement
with hazardous materials incidents unless otherwise directed by federal or state statutes, rules
and/or regulations. [WVC 29-3-9(i)]
Unlike requirements for industrial hazardous materials incident responders, this Fire Commission
requirement does not specifically address recurrent training. All active industrial hazardous materials
54 The criteria are located in WVR 87-6-5.
55 The administrative policy is found at WVC 29-3-9(i).
56 NFPA 472 was renamed the Standard for Competence of Responders to Hazardous Materials/Weapons of Mass
Destruction Incidents in the 2008 revision.
Little General Store September 2008
49
responders in the United States are required to undergo an annual refresher course. Yet firefighters, who
are likely to respond to many incidents involving gasoline, natural gas, or propane—all hazardous
materials—in any given year, are not specifically required by the West Virginia Code to complete an
annual refresher course or perform regular response drills. As discussed in Section 5.4, the IC in the
Ghent incident had not trained for or practiced hazardous materials emergency response in almost 10
years.
6.3.2 West Virginia Fire Code
West Virginia has a comprehensive statewide fire code.57 The fire code incorporates, by reference, the
majority of the NFPA’s standards and guides. Incorporated standards include NFPA 472, which addresses
response to hazardous materials incidents. NFPA 472 defines the levels of response to hazardous
materials incidents—awareness, operations, technician, incident commander, and specialist employee—
and outlines basic expectations for responders at each level.
The NFPA’s definition of awareness level personnel is consistent with OSHA’s definition:58 awareness
level personnel are those who, during the course of their normal job functions, could encounter
emergencies involving hazardous materials. These persons must be able to recognize hazardous materials
emergencies, protect themselves, call for trained assistance, and secure the area.59 Awareness level would
apply to propane service technicians like the junior technician at Little General when the release began,
since they may encounter propane emergencies during the course of their duties.
Operations level personnel respond to already-discovered hazardous materials incidents, taking actions to
protect people, the environment, and property from the effects of the release. They are qualified to take
57 West Virginia Rule Title 87 Series 1 Fire Code.
58 29 CFR 1910.120
59 NFPA 472 Chapter 4
Little General Store September 2008
50
defensive actions to mitigate the damage caused by a hazardous materials incident, but not to actively
handle or attempt to contain the hazardous material(s). This type of action might include assessing the
scene and planning a response, establishing evacuation perimeters, setting up communications, and
initiating the incident command system. The operations level designation applies to anyone arriving at
the scene in response to a hazardous materials emergency; in Ghent, the lead technician, the EMTs, and
the fire department personnel acted at an operations level by arriving at the Little General in response to
the junior technician’s phone call.
Hazardous materials technicians, in contrast, are those qualified to take defensive and offensive actions in
response to an incident. Offensive actions entail entering the incident “hot zone” to attempt to control a
release, and require additional training beyond the operations level.60 Additionally, hazardous materials
responders may seek the help of specialists: industry employees familiar with and trained in the hazards
and systems of a given material, like propane. Specialist employees must also meet specific NFPA 472
competencies to be qualified to perform release control actions. Propane service employees who assist
emergency responders physically secure releases must meet at least these competencies. Although the
evidence and witness statements indicate that the Appalachian Heating employees and Ghent Volunteer
Fire Department personnel were standing in the vapor cloud and attempting stop the release, none were
trained to the technician or specialist level.
60 Technician level training corresponds to 29 CFR 1910.120 40-hour HAZWOPER training.
Little General Store September 2008
51
7.0 Findings
1. The propane service technicians, emergency responders, and store employees did not evacuate the
area as recommended by nationally accepted guidance for propane emergencies. In fact, emergency
responders and the propane service technicians were observed standing in the propane vapor cloud.
2. Propane companies, the National Propane Gas Association, and the U.S. Department of
Transportation recommend evacuation to a safe distance as the first action in response to a propane
release.
3. The fire department captain, the propane service technicians, and the Little General Store employees
were unaware of the dangerous propane accumulation inside the building.
4. A defect in the existing tank’s liquid withdrawal valve caused it to malfunction and remain in an open
position.
5. About 25 minutes elapsed from the time the release began until the explosion.
6. Both the U.S. Occupational Safety and Health Administration’s and the West Virginia Fire
Commission’s propane standards require a distance of at least 10 feet between 500-gallon propane
tanks and buildings such as the Little General Store. However, when the Southern Sun propane
company installed the propane tank in 1994 it was placed directly against the Little General Store’s
exterior back wall.
7. Ferrellgas, which acquired Southern Sun in 1996, allowed the tank to remain against the building’s
exterior wall for more than 10 years.
8. The placement of the 500-gallon propane tank against the building’s exterior wall provided propane a
direct pathway into the store’s interior during the release.
9. Ferrellgas management’s quality review program functions as a basic safety inspection rather than a
management systems audit.
Little General Store September 2008
52
10. The junior propane service technician who was servicing to the tank on the day of the incident had no
formal training and did not recognize the defect in the withdrawal valve. He was also working alone
even though he had been on the job for only one and a half months.
11. Propane service technicians commonly do not receive emergency response training.
12. The propane industry’s primary training curriculum (the Certified Employee Training Program)
consists of procedures and materials for performing routine (non-emergency) tasks only.
13. The Occupational Safety and Health Administration’s and National Fire Protection Association’s
propane standards require training but do not include curricula, practical exercises, emergency
actions, or knowledge evaluation tools.
14. A propane emergency significant enough for fire department response is reported nearly everyday in
the United States. Only gasoline and natural gas are involved in more hazardous materials
emergencies.
15. 911 operators in the United States lack propane emergency guidance to help them collect important
information from callers, offer life-saving advice, and convey relevant information to first responders.
16. Minimal information on the nature of the incident was exchanged between the propane service
technician and the 911 operator. Therefore, the only information the 911 operator provided the Ghent
Volunteer Fire Department responders was the incident location and the “report of a propane leak.”
17. Firefighters in West Virginia are required to attend a minimum of four hours of hazardous materials
emergency response training as part of their initial training sequence but refresher training is not
required. The responding Ghent Volunteer Fire Department captain last attended a hazardous
materials response course in 1998.
18. Propane safety and emergency training is voluntary for fire department personnel in West Virginia.
None of the responders from the Ghent Volunteer Fire Department had specific training relating to
propane emergencies.
Little General Store September 2008
53
8.0 Causes
1. The Ferrellgas inspection and audit program did not identify the tank location as a hazard.
Consequently, the tank remained against the building for more than 10 years.
2. Appalachian Heating did not formally train the junior technician, and on the day of the incident he
was working alone.
3. Emergency responders were not trained to recognize the need for immediate evacuation during liquid
propane releases.
Little General Store September 2008
54
9.0 Recommendations
Governor and Legislature of the State of West Virginia
2007-04-I-WV-R1
Require training and qualification of individuals who operate bulk propane plants, dispense and deliver
propane, install and service propane systems, and install propane appliances. The training and
qualification requirements should be comparable to those of existing propane industry programs such as
the Certified Employee Training Program.
West Virginia Fire Commission
2007-04-I-WV-R2
Revise the Fire Commission rules and codes to require annual hazardous materials response refresher
training for all firefighters in West Virginia.
2007-04-I-WV-R3
Revise the Fire Commission rules and codes to require that all West Virginia fire departments perform at
least one hazardous material response drill annually.
West Virginia Office of Emergency Medical Services
2007-04-I-WV-R4
Revise the Office of Emergency Medical Services rules and codes to require annual hazardous materials
response refresher training for all emergency medical personnel in West Virginia.
Little General Store September 2008
55
National Fire Protection Association
2007-04-I-WV-R5
In the Liquefied Petroleum Gas Code (NFPA 58) “Qualifications for Personnel” section, specify training
requirements (including supervised on-the-job training), training curricula, competencies, and testing
through written examination and performance evaluation, or reference a nationally recognized curriculum
for these requirements.
Association of Public-Safety Communications Officials
2007-04-I-WV-R6
Develop a guide card for propane emergencies to assist 911 operators in the collection of pertinent
information on propane emergencies. The questionnaire in Section 1.9.1 in the Propane Education and
Research Council’s Certified Employee Training Program may be used as a model.
Propane Education and Research Council
2007-04-I-WV-R7
Revise the Certified Employee Training Program to include
• Procedures for transfer of liquid propane from tank to tank, or
• The prohibition of the transfer of liquid propane from tank to tank.
2007-04-I-WV-R8
Revise the Certified Employee Training Program to include emergency response guidance for propane
service technicians who respond to propane emergencies similar to guidance provided to emergency
responders in the Propane Emergencies program.
Little General Store September 2008
56
National Propane Gas Association
2007-04-I-WV-R9
Submit a request to the United States Occupational Safety and Health Administration for a letter of
interpretation to determine if the Certified Employee Training Program curriculum meets the training
requirements in 29 CFR 1910.110.
2007-04-I-WV-R10
Work with the West Virginia E911 Council with development of propane emergency guidance by
providing the Council with the customer leak questionnaire located in Section 1.9.1 of the Certified
Employee Training Program and technical assistance.
West Virginia E911 Council
2007-04-I-WV-R11
Work with the National Propane Gas Association to develop and distribute propane emergency guidance
for use by all county and municipal 911 communication centers in West Virginia.
Little General Store September 2008
57
Ferrellgas
2007-04-I-WV-R12
Establish and implement a comprehensive safety management system that includes at a minimum:
• An auditing program developed in accordance with generally accepted methodologies to
monitor the performance and effectiveness of safety management systems and personnel
at all levels,
• An inspection program that uses NFPA 58 as a guide to systematically inspect all
customer propane systems and identify all deficiencies,
• A means of tracking audits and inspections and identified deficiencies,
• A means of tracking corrective actions,
• A means of collecting and using audit and inspection data for trend analysis and
organizational learning,
• A means of periodically reporting audit and inspection trends to the Board of Directors
and Managing Board, and
• A provision for periodic safety management system audits conducted by a third party
competent in the requirements of NFPA 58.
Little General Store September 2008
58
By the
U.S. Chemical Safety and Hazard Investigation Board
John S. Bresland
Chair
Gary Visscher
Member
William Wark
Member
William Wright
Member
Date of Board Approval
Little General Store September 2008
59
References
Center for Chemical Process Safety (CCPS), 1993. Guidelines for Auditing Process Safety Management
Systems, American Institute of Chemical Engineers (AIChE).
Ferrellgas Partners, L.P., 2006. US SEC Form 10-K – Annual Report Pursuant to Section 13 or 15(d) of
the Securities Exchange Act of 1934.
Hildebrand, M.S., and G. G. Noll, 2007. Propane Emergencies,3rd ed., National Propane Gas Association
(NPGA) and Propane Education & Research Council (PERC).
Lemoff, T. C., ed., 1998. LP-Gas Code Handbook (5th edition), National Fire Protection Association
(NFPA).
National Fire Protection Association (NFPA), 2008a. Liquefied Petroleum Gas Code, NFPA 58.
NFPA, 2008b. National Fuel Gas Code, NFPA 54.
NFPA, 2008c. Standard for Competence of Responders to Hazardous Materials/Weapons of Mass
Destruction Incidents, NFPA 472.
Occupational Safety and Health Administration (OSHA), 2007. Storage and Handling of Liquefied
Petroleum Gases, 29 CFR 1910.110, OSHA.
Petersen, D., 1996. Analyzing Safety System Effectiveness, Third Edition, John Wiley & Sons, Inc.
Raleigh County Emergency Services Agency, 2006. Raleigh County Emergency Operations Plan (EOP).
West Virginia Fire Commission, 2004. State Fire Code, Title 87 Series 1 (87CSR1), West Virginia
Legislative Rules.
Weinstein, M.B., 1997. Total Quality Safety Management and Auditing, Lewis Publishers.
West Virginia Legislature, 2007. Fire Prevention and Control Act, Chapter 29 Article 3 (WVR 29-3),
West Virginia Code.
Little General Store September 2008
60
Appendix A
Ferrellgas Installation Reviews of the Incident Tank
Little General Store September 2008
61
Little General Store September 2008
62
Little General Store September 2008
63
Appendix B
911 Call and Initial Fire Department Dispatch
Little General Store September 2008
64
Emergency 911 Call – Report of Release
AUTOMATED VOICE: Conversation recorded on January 30th, 2007, at 10:40, on Channel 7.
DISPATCHER: 9-1-1, where is your emergency?
CALLER: Yes, ma’am. I need to — I need a fire department down at the Little General Store in Ghent. I
need a — I got a propane leak — I need their help to secure the area.
DISPATCHER: A propane leak?
CALLER: Yes, ma’am.
DISPATCHER: What’s the address? Is that the one on Odd Road?
CALLER: No, ma’am. It’s the one actually on, uh — the one going towards Shady Springs. With the
Godfather’s Pizza.
DISPATCHER: Do you know the address?
CALLER: No, I don’t, ma’am. Right in front of Flat Top Lake.
DISPATCHER: Flat Top Lake.
CALLER: Yes, ma’am.
DISPATCHER: And this is Little General, right?
CALLER: Yes, ma’am.
DISPATCHER: And this is a propane leak? What is your name?
Little General Store September 2008
65
CALLER: XXXXXXXXXXXXX;1 I work for Appalachian Heating; we’ve had a dysfunction in the
tanks, and I have a leaky tank.
DISPATCHER: Okay. All right, I’ll get the fire department out there for you.
CALLER: Thank you. Bye-bye.
DISPATCHER: Thank you.
Emergency Dispatch – Initial Dispatch Notification to Ghent
Volunteer Fire Department
AUTOMATED VOICE: Conversation recorded on January 30th, 2007, at 10:43, on Channel 21. (Fire
tones.)
DISPATCHER: Raleigh Control, Station 110 — Station 110 — you need to respond to Flat Top Road in
Ghent, the Little General, across from Flat Top Lake; report of a propane leak. Repeating, Station 110,
need units to respond to Flat Top Road at the Little General, report of a propane leak. WGC808, 10:44.
1 Name removed.
Little General Store September 2008
66
Appendix C
Recent Propane Incidents
Little General Store September 2008
67
Recent Propane Incidents
CSB identified the following incidents for further research based on their similarity to the Little General
incident. The CSB interviewed the propane companies, emergency responders, and local investigators
involved in these incidents to collect relevant information. The CSB did not conduct an independent
investigation.
1.0 Propane Release Incidents
1.1 Aberdeen, Washington
1.1.1 Incident Description
On October 23, 2007, the Southshore Mall in Aberdeen, Washington, was evacuated due to a vapor
release from a 1,150-gallon propane tank. The tank was located in a utility yard approximately 25 feet
from the mall restaurants for which it supplied heat and fuel. Shortly after 2:00 pm, a mall employee
heard a “pop” and, upon going outside to investigate, discovered the leaking tank. The fire department
responded, evacuated the mall, and isolated the scene.
The fire department called the propane company that owned the tank for incident support. The company
sent one service technician to assist the fire department. The responding technician found the release was
from the tank’s fill valve, which had become stuck partially open. The technician installed a double-
check valve on the leaking fill valve to stop the release. The fire department lifted the evacuation order
by 2:45 pm.
1.1.2 Incident Responders
The propane company involved in this incident reported having a preexisting working relationship with
the fire department. All the propane company’s employees train on job procedures with CETP and
Little General Store September 2008
68
practice emergency response scenarios. The service technician who was sent to assist the fire department
had both CETP training on all propane service and delivery procedures and on-the-job experience as a
service technician and a delivery driver.
In emergency response cases like this one, the fire department is in charge of the scene, with the gas
company employees on the scene to assist and advise the fire department as necessary. Occasionally,
propane service technicians responding at the request of a fire department take offensive steps to stop a
release—in this case, by installing a valve. The fire department’s decision to allow the propane service
technician to repair the tank was different from the Ghent incident in several important ways:
• The fire department had evacuated and secured the scene prior to the technician’s entry.
• The vapor release from the leaking fill valve was much less severe than the liquid propane release
in Ghent, requiring only a small repair for the tank to remain in service safely.
• The technician was CETP-certified and experienced, had drilled on emergency response
scenarios, and had previously worked with the fire department on propane incidents.
When emergency responders and service technicians have clearly established and rehearsed roles in an
incident command structure, both groups are better equipped to implement their training.
1.2 Lynchburg, Virginia
1.2.1 Incident Description
At about 1:00 pm on November 26, 2007, a traffic accident caused a vehicle to crash into a 1,000-gallon
propane tank used for filling cylinders, located outside of a True Value Hardware store in Lynchburg,
Virginia. When the vehicle struck the tank, the inch-and-a-half liquid line connecting a pump and the
tank was severed. The tank settled on top of the severed pipe, which in turn prevented the excess-flow
valve from seating, resulting in a liquid propane release. Hardware store employees called 911 and the
Little General Store September 2008
69
propane company that owned the tank. Although the 911 operator dispatched the hazardous materials
response team to the incident as a gas leak, a citizen driving past called the team to clarify that it was a
propane release.
The team arrived and secured the area, evacuating shoppers from the rear of the hardware store on foot.
Since the valve could not be repaired, the response team used water spray to disperse the vapor cloud
formed by the liquid release, while keeping the area evacuated. Once most of the propane in the tank had
been released, the response team approached the tank and used a strap to completely seal it. They then
loaded the sealed tank onto a truck for removal from the area. The release lasted two hours.
Propane service technicians from the company that owned the tank also responded. They arrived after the
hazardous materials response team, and attempted to approach the tank. Response team members
removed the technicians to the incident command post, outside of the release area.
1.2.2 Incident Responders
The responding hazardous materials team is an all-career unit, with all members trained to technician
level (40-hour) training, which qualifies them to take offensive action in response to releases. According
to responders, propane incidents occur frequently in the area; although most involve residential propane
cylinders, the hazmat team has responded to several major incidents. As a result of the prevalence of
propane in the area, its hazards are a focus in training sessions. With this training, responders were able
to quickly assess the uncontrolled release and evacuate the area.
All propane service technicians at the company that owned the tank, a small locally owned business,
receive CETP training. Additionally, the company has trained with the fire department in CETP
procedures to help emergency responders understand propane systems. However, the CETP program has
no emergency response component, and propane service technicians do not train with responders the
incident command system or other aspects of hazmat response. As in Ghent, propane service technicians
Little General Store September 2008
70
attempted to take an active role in release mitigation without hazardous materials or emergency response
training. However, unlike Ghent, fire department responders in this incident were trained and
experienced in the hazards of propane releases, and evacuated everyone, including technicians, from the
area.
1.3 Bristow, Virginia
1.3.1 Incident Description
On May 19, 2008, the Linton Hall School in Bristow, Virginia, was evacuated due to a vapor release from
a 500-gallon propane tank. The tank was located 15 to 20 feet from the school, near the cafeteria kitchen.
At about 11:20 am, a trash truck backed into a chain connecting the tank to a light pole,
1 overturning the tank and shearing off its service valve, causing a propane vapor release. School
administrators called 911 and the propane company that owned the tank, and evacuated the 205 students
and 30 faculty members according to routine fire drill procedures.
The 911 operator who received the call dispatched the county fire department to the incident as an outside
gas leak. The first responding fire engine company arrived on scene, assessed the situation, and revised
the call to an inside gas leak due to the tank’s proximity to the school kitchen and water heater,
dispatching the hazardous materials response team.2 When the team arrived, it sprayed water in a fog
pattern to push released vapor down and away from the school, then used a wooden dowel and a towel to
plug the leak. Propane service technicians from the company that owns the tank arrived after the team
had secured the release and moved the tank to a large open parking lot away from the school to repair the
1 A “No Parking” sign hung from the chain.
2 The dispatching system in this county requires the hazardous materials unit to respond to all interior gas leaks.
Little General Store September 2008
71
valve. No one was injured. Students remained at a church across the street until parents arrived to take
them home.
1.3.2 Incident Responders
The hazardous materials team involved in this incident has responded to several propane incidents in the
last year, all of which have been vapor releases. They receive technical information from propane service
technicians, but take release mitigation actions themselves. To protect propane service technicians from
entering the release area, the team uses methods including
• taking digital photographs of tank damage so that technicians can see and assess the damage on a
remote laptop computer,
• bringing similar tanks or other equipment to the scene so that technicians can point out features
the team will find on the incident tank, and
• escorting technicians to the edge of the secured release area to point out potential leak sources for
team members.
Additionally, the hazardous materials response team trains local 911 operators to use follow-up questions
to elicit needed information from callers reporting hazardous materials incidents. In this incident, as in
Ghent, the 911 operator conveyed little information to the fire department. However, unlike Ghent
responders, the first fire engine company to arrive immediately recognized the danger posed by the
release’s proximity to a building and called for hazardous materials backup.
The propane service technicians responding to this incident were branch employees of a national propane
company that uses CETP. In the event of a propane emergency, whichever technicians are working
nearest the incident stop work immediately and respond. Since these responders are all trained with
CETP, they are familiar with what types of incidents they can mitigate using normal propane procedures.
Little General Store September 2008
72
In this case, propane service technicians attempted to repair the tank only after the fire department had
secured it and removed it from nearby buildings.
2.0 Propane Release and Fire Incidents
2.1 Sallis, Mississippi
2.1.1 Incident Description
On May 13, 2007, a propane explosion occurred at the Longhorn’s Steak House in Sallis, Mississippi,
while the restaurant was closed and two propane service technicians were transferring propane from an
old 500-gallon tank into a new 500-gallon tank behind the building. The transfer procedure had been
completed and the new tank was leak-tested and placed in service. The technicians next planned to
remove the old tank (which still contained some propane) from the premises, but were unable to load it
onto the service truck. They left and returned later in the day with a trailer. As the technicians were
loading the old tank onto the trailer, it slid to one side, severing a valve at the bottom of the tank. The
technicians heard a loud noise and saw liquid propane releasing from the bottom of the tank and forming
a vapor cloud along the ground. They immediately ran toward the front of the building; the propane
reached an unknown ignition source and a flash fire occurred seconds later. Another person who was
near the front door of the restaurant at the time was badly burned in the flash fire. The building and a
pickup truck parked in front were destroyed. Local volunteer fire departments responded, and the
Mississippi State Fire Marshal’s office investigated.
2.1.2 Incident Responders
The propane service technicians were from a local propane company. Although they had already
completed a tank-to-tank transfer without incident, they immediately realized the danger of a liquid
propane release from the damaged tank. Since the propane reached an ignition source only seconds after
Little General Store September 2008
73
the release began, the technicians were unable to evacuate the person in front of the restaurant before the
flash fire. Both reported to investigators that they had attempted to do so.
Local firefighters responded and reported the incident as a structure fire. The county sheriff’s department
informed the State Fire Marshal’s Office that propane was involved, and LP Gas Division investigators
completed the state’s report.
2.2 Danville, Alabama
2.2.1 Incident Description
On May 20, 2008, three propane service technicians were attempting to perform a tank-to-tank transfer of
liquid propane at a farm in Danville, Alabama, when a propane vapor release caused a flash fire. A
1,000-gallon propane tank at a chicken house was reported by property owners to be leaking around the
pressure relief valve. Technicians were to replace the half-full tank with a new tank and transfer the
remaining propane.
The technicians removed the leaking tank from its installed location and placed it to the side, then set the
new tank in its place. Employees next used the service truck boom to suspend the leaking tank upside
down and at an angle, near the new tank. They planned to use the filler valve to evacuate the old tank,
hooking it to a hose with a filler valve adaptor. However, damage to the filler valve caused it to stick in
an open position and release propane vapor. The vapor reached the service truck, idling near the transfer
location, and ignited setting fire to the chicken house.
The town’s volunteer fire department was dispatched to a structure fire. When firefighters arrived, they
discovered that the fire was due to a propane release. The firefighters then moved the three injured
technicians away from the fire, used water to cool the propane tank, and approached the tank in full
protective equipment to attempt to close the valve. They discovered that the valve had been sheared off
entirely and withdrew, continuing to contain the fire to prevent it from reaching other propane tanks
Little General Store September 2008
74
located 25 to 30 feet away. The firefighters called for the nearest hazardous materials response team to
assist them in using thermal imaging to determine how much propane remained in the tanks. The incident
tank was empty; no other tanks released. The chicken house was destroyed. All three propane service
technicians were severely burned; one died in hospital.
2.2.2 Incident Responders
The volunteer firefighters in Danville are all trained using the Propane Emergencies curriculum, which is
offered through the state fire college as a two-day program. The fire chief requires that they complete the
program twice before responding to propane incidents. This training includes practice on controlled
propane releases and fires, and allows the fire department to mitigate propane emergencies without
relying on propane companies. If firefighters cannot contain a release, they may let propane service
technicians work on the tank, but only if no fire is involved and the area has been secured.
Little General Store September 2008
75
Appendix D
Table of State Training, Testing, and Licensing Requirements
for Propane Service Technicians
Little General Store September 2008
76
For Propane Service
Technicians, this state
A
rk
an
sa
s
A
la
ba
m
a
C
ol
or
ad
o
Fl
or
id
a
Io
w
a
M
ai
ne
M
ar
yl
an
d
M
ic
hi
ga
n
M
is
so
ur
i
N
eb
ra
sk
a
N
ew
J
er
se
y
O
kl
ah
om
a
T
ex
as
V
er
m
on
t
Requires training x x x x x x x x x x
Requires testing x x x x x x x x x
Requires recurrent
training x* x x x x x* x
Accepts CETP x x x x x x x
Requires CETP† x† x x† x x x
Requires License x x x
* Only safety training required on a recurrent basis
† CETP or equivalent training required
Little General Store September 2008
77
Appendix E
United States EPA – 40 CFR 311
Little General Store September 2008
78
311.1 Scope and application.
The substantive provisions found at 29 CFR 1910.120 on and after March 6, 1990, and before March 6,
1990, found at 54 FR 9317 (March 6, 1989), apply to State and local government employees engaged in
hazardous waste operations, as defined in 29 CFR 1910.120(a), in States that do not have a State plan
approved under section 18 of the Occupational Safety and Health Act of 1970.
311.2 Definition of employee.
Employee in Sec. 311.1 is defined as a compensated or noncompensated worker who is controlled
directly by a State or local government, as contrasted to an independent contractor.
- copy for updated table of contents
- draft for pages through 6
- copy for pages 7 through 12
- copy for pages 13 through end
- copy for pages appendix headings
CHAPTER11
p.;
Cause and Effect Analysis
The cause and effect analysis is a structured approach to determine th
e
causes and effects of events, which lead to the accident. There are many types
of cause and effect analysis used in accident investigations, and this book
will focus on the Apollo method (cause and effect process) and the Ishikawa
Fishbone analysis ( developed by Kaoru Ishikawa) that the best way to view the
causes and effects was to draw it out like a fish skeleton, with the problem at
the head of the fish and the bones representing the causes. Another view of
a cause and effect analysis is the “5 Whys” or “Questions to the Void.” The
5 Whys approach is to analyze the event to five levels (give or take a couple)
until it is clear that the root cause is found.
The cause and effect approach is a structured approach, and examining
the Apollo Method is based on (1) defining the problem; (2) developing a
causal understanding of why the problem occurred (cause and effect chart);
(3) identifying solutions; and ( 4) implementing and monitoring effectiveness
of the best solutions (Gano 2007). For accident investigations, the problem
has been identified, so for this approach, the only part discussed will be the
cause and effect approach and the cause and effect chart.
117
P11rt Ill: Anab•timl Trchniquu
11 8
Cause and Effect Approach
The cause and effect analysis is best performed as a charc .
events and causal factors analysis or tree analysis. The k s1Illilar to h
. . ey to the t e
effect analysts 1s che thought process of developing the effe caus, •nd
and repeating that sequence. The process is the sam h cts of cau
. . e w ether . ses Apollo Cause and Effect, Ishikawa F1shbone, or the 5 Wh using the ·th h ·d d ys analy · approach is to start WI t e aca ent an analyze until • sis. ‘l-1
1t reache th ‘n,
factors . The approoch to this an alysis can be perfo d s e caus,i
rme on c software (Apollo 2011 ) or done on a sheet of paper h . 0 rnputer
. , w lteboard using adhesive nores. The benefit of a thorough cau , or bi·
. . se and effe . keep analyzing unnl all causal fac tors are found includin ct IS to
h · th · . g any sysr . causes, althoug many runes . e 1nvesugato r will stop at th enuc
causes. e superficial
The approach to these three cause and effect analyses .
. are very Similar
Exlubit 11.1 displays the charnng structure for the three approaches. ·
The Apollo Method is a software-based m ethod that leads th • .
e tnvesngator
through the approach with an effect followed by at least two causes. Each
effect is “’caused by” a conditional cause and an action cause. According
ro the Apollo method, “every time we ask ‘why,’ we should find at leasr
ru-o causes and for each of these causes we should find at least two more
causes resulting in four causes, and so on” (Gano 2008). Normally the
chan is developed from left to right.
The fishbone diagram has been used for many years in business and as a
quality rool for manufacturing. The diagram resembles the skeleton of a
fish and focuses on the causes rather than the symptoms of a problem.
The problem statement is the head of the fish along the fish’s backbone.
The ne..u step is the brainstorming or in case of an accident investigation, the
facrs and analysis are analyzed to structure the big bones of the fish that
are connected to the backbone (Ishikawa 1968). What areas of the acciden<
investigation are the major types of causes? Some investigators _use the
same major big bones such as people, policies, procedures, eqwpmem,
and measurement, materials, and/ or environme nt. Others change each
Chapter 11: Cause and Effect Anafysis
es of causes for the accident, such as lockout
the relevant typ . h Thi
” “””n ID . . procedure and management overstg t. s wa5•- dure uammg, ‘ . h
.,gou• proc~ ‘m different and more specific to the accident; owever,
-,kes each agra_ ti ier to start the process. Most fish bone diagrams
1…- , is a lit e eas . k
,)le firS t wa) d ent on the right of the page and will actually wor
.th the hea or ev h th fi h’ ,..,. wt 1 f . however it really doesn’t matter whic way e s s bacJ.,,ar
he<
15
. thod of solving a problem or finding the causes of
5 Whys IS a me . • The t dly asking-at least five tunes-why the problem, ·dent by repea e
an ,co .d t occurred and then why that cause occurred to explore
. oracoen . . h
esenr, d effect relationship and discover the causal factors (Haig t
the cause an . f th h thi .
X’h,ile this is not a precise techruque as many o e ot ers, s IS an
2008). ·ct · · · h the effective technique if the investigator or acct ent mvest:1gatton ~earn as
. th faces and analyses of the accident. This is not a technique you use
m-
l'”rt Ill· . A 11aly1fr,,/ T , .
8CfJl1Jq11e.s
120
Fishbone Diagram
~eople J
~ _l!kof
~~ures
1~–~ // Equipment ] [ Policies J
SWhys
Event
Why?
Why? Bad
Housekeeping
Chapter 11: Ca11st and Ejftd Ana!Jsis
cause and Effect Process
· to deterrnine the effect (event, accident, problem). In an
· ·· Thi . .
1n’ . ·gacion the effect is the ace1dent, UlJury, or damage. s 1s
•dent lflvesu , . ‘”‘ f’ er that effect is then analyzed by asking: •’Why?”; •’What was this
,hefirSte ,e ‘ , • · ., b ,, … or ”\VhY did this happen?’ to get to the next level. This 1s then
ted until the causal factors are found. One of the issues with any causal
c,,11SCU } · ‘
::~ analysis is knowing when to stop. If there is lack of investigating or
1:n,wledge of the facts and analysis of the accident, then these techniques
,ill not lead to the causal factors but will stop short and lead to symptoms
of problems instead. Tbere are many types of problems that can occur with
the cause and effect analysis, as listed in Exhibit 11.2.
Exhibit 11.2
;:::: PROBLEMS WITH,.c;AUSE Afilf EFFECT
• Stopping too soon (stopping at th
supert1c1al causes) e
• The need to place blame
• Not having enou h of the accident o;st~~owledge (facts)
analysis too soon ,ng the
• Not looking at all issues
(management pol’
supervision, t/aini~cy’
human factors des~g• ‘ n, etc.)
Exarnpl S Tu, e cenario
example scenarj
]
P
)~cal techr,j nt investiga . ould be f . l e other techni
ques to find th Uons there is b’ airy straightfo ques
e causal facto a tg benefit of _rward. This
rs. If using the 5 ~sing multiple
ys techni que,
121
Parl II/: Analytital I,,bniqueJ
122
the accident would be laid out similar to the tree analy ·
. s15 and WouJd shape. For purposes of usrng the technique, let’s do take that
a couple of . . of the 5 Whys. The top event would be the individual fallin ueranons
\Vhy did the individual fall off the ladder> Exhibit l l 3 di off th• ladder.
th 5 Wh h . w n.. . thi . sp ays the s••-e ys tee ruque. w uen usrng s technique, you will h ~, of
many iterations to get all of the causal factors . ave to per/onn
Exhibit 11.3
Event Falling off
ladder
.0.
Why? Ladder hit
by forklift
.0.
Why? Failure to
barricade aisle
Summary
There are three types of cause and effect analysis: Apollo Method, fishbone
diagram, and the 5 Whys. Each of these techniques is a structured approachw
stan with the accident and analyze it until it reaches the causal factors. While
there are problems with this technique, if it is used by experienced trruned
~ ve5rigators, these techniques are very effective at reaching causal factors,
mcluding systemic causes.
*
Chapter 11: Couse a11d Ejftrl A11olysis
REVIEW QUESTIONS
d effect analysis?
es of cause an . ,
ethe three !’/P and effect analysts .
1.wttatar . three types of cause
he differences tn the
iWhatare t ctured approach?
J. wt,at is the Apallo method stru . fusing a cause and effect analysis
blems orfatlures o
4 W,,ataresome of the pro
. such as the 5 Whys? . does the investigator
d ffect what questions
5. When analyzing a cause an e ‘
continually ask? A llo method and the
I the example scenario with the po 6. Continue to ana yze
fishbo ne diagram.
CHAPTER12
I ••r
Specialized and
Computerized Techniques
So far this book has discussed five major types of analytical techniques:
events and causal factors analysis, change analysis, barrier analysis, tree analysis,
and cause and effect analysis. Many other specialized analytical techniques can
be used in accident investigations, and each has a role in certain situations.
(NOTE: 1n some types of accidents, these specialized techniques may yield so
much information and so many possible causes that the investigator cannot
deal with them. The process becomes too confusing and frustrating. The
mvestigator must learn to use the appropriate technique for each type of
accident)
Some of the newest analytical techniques for accident investigation involve
~-~oo d . all . . mputers an software. Some programs s!Illply ow investigators
to portray a ·a · ·d cc, ents graphically, but others actually help to analyze acc1 ents.
h ” . d . e tee ruques range from computerized trees to fully an!Illate
accident re . construction programs.
125
Part lll: At1u!) t1l”al Trrhm
126
Specialized Techniques
Time Loss Analysis
Time loss_ analysis ,~•as developed for the National Tran
Board. It 1s a graphical analysis tool that investigators casportation Sifett
ro ~nderstand, de~’elop, an~ evaluate interventions and em: use proacti\”e:\;.
~coons, ~nd r~acav~ly co give credit to good emergenc , re tgency responst
1mervent10ns 1n accident situations. ) sponse effons or
Time loss analysis helps the investigator to evaluate how the _ .
em~ncy response or loss control actions affected the loss cau llnung of
ac~dent and to analyze _the losses that occurred as the events lea . sed hr the
acadeot progressed. It 1s useful in determining how 1 up to the oss control mterv .
changed (or could have changed) the amount of loss a d h . . enao~
n owumeincrcasea
~r decre~sed lo~ses ~r costs. It provides a way to analyze the interventions
l.Il an acadent s1 ruat1on and determine how they changed the course of the
accident sequence (SSDC 1987) .
To use this technique, investigators follow these seeps:
Discover and analyze all interventions that took place leading up to
and during the accident.
Determine whether these interventions increased, decreased, or had
no effect on the outcome (the accident sequence). Forexample,ini
car crash with an injury, what is the effect o f the vehicle having anti-lock
brakes? D o the brakes allow the car to sto p any sooner or in a shorter
distance? What about airbags? What about crumple zones?
Assign a time value to each intervention. The time value can be an acrw.l
number or a relative positio n on the chart. For example, the brakes stan
to act as soon as the driver senses danger and applies them. The anti-Jock
brake mechanism activates when the wheels begin to lock. The air b~
activate at the moment of impact. The crumple zones crumple after che
. . h damage to che
initial impact as the car continues to m ove, increasmg t e
car but distributing energy away from the passengers.
2!’.iiW
S
. f-,,ed and Computen’z.rd Tedmiqurs
Chapter 12: p,na1,
k d in dollars or units of
The loss axis is usually mar e b er of people injured ,
e the loss- d d or gas leaked, num
‘ p,.,i112.t ount of water £loo e
•• (iJll f . hides damaged). . c the simple car crash
I# o ,e al •sts chart ior .
n1,l!ll . wires a time loss an ) . fall three intervenoo ns-
£.tlubit 12.1 l~ lowest amount ofloss occurs t ed. Even if all three
b(
,h(·lflo-lOC ill be some damage; however,
¢ therew
i,’t , ,ill be much greater.
oi\oSSW
Analysis 11uman Factors al . and a human factors or
ri f h an factors an ysts,
1’1..- ire man}’ methods o um al ‘ The basis of human factors
.. ~- h uld erform the an l ses. .
..oonomics expert s o p chin / w· orunent interaction and to detemune .,- . tify human/ ma e en f
~’Sisis to1den : h d an effect on the accident. There is som e type o
~ticther the intw.cnon a f cident The key is to determine the
involvement in all aspects o an ac . . ~= involvement and the human capabilities to perform the task.
Exhibit 12.1
(
+
TIME
LOSS ANALYSIS
TO – Driver senses danger
I _______ T 1 -Driver applies brakes \ I T 2 -Anti-lock brakes deploy T 3 – Airbags deploy I, T 4 -Crumple zones crumple
t
\~~ Natural : : : : , course of
: \ \ \ \ \ accident
9 : \ : Courseof
\ : ; \ l ~i~~d=l~t
: \ interventions
-~ ineffect
To T, T2 T3 T4
TIME
128
To ~ nduct n human/ machine / environment analysis
c.,-pc:.n tollm.vs these steps: of an accid
•n~ th
Analyze ho·w the human interacted with th ‘
e rnachin environment, etc. e, eqUiPrnen
Llst the bad interactions in the accident t,
– – sequence l
mter.1cao ns that do not favor the capabilities of the h · hese are th
capabilities include physical interactions (stre h umf! a~ body. I-I1.1rn ‘
. . . . ngt , ex,bi]j an
m~a~ n, eyesight, ~eann~, men~al tnteractions (knowled e ~• ra~ge of
tr.11n1Ilg), and emononal mteracnons (morale mon· . g • lrltelligence
, vanon, attitud •
Other types of human factors analysis analyze anthropome . e).
rk h · I cl · d · · try, btomechani ,vo – p ysio ogy, s ecnon an traU11ng of personnel, job tasks and w ts,
In this of analysis, the focus i~ on the work environme~t that orkloads.
bad beha”‘,or. The safety professional tries to eliminat th b Produces
. . eeadbeh · instead of focusing on human error (Oakley and Smith 2000). avior
Integrated Accident Event Matrix
An integrated accident event matrb. includes a list of all individuals who
were at the scene of an accident and a time-based chart that shows their
interactions (DOE 1999). This matrix allows an investigator to analyze what
each individual was doing at the time of the accident. Although the chan can
include any amount of time before the accident, it usually only covers the ten
co twenty minutes before the accident occurred. This type of matrix is helpful
for many types of accident investigations and is very simple to perform. It is
a low-tech method of accident reconstruction.
To perform an integrated accid ent event matrix, the investigator follows
these steps:
List all of the individuals who may hav e been involved in the accident in
the left column of the matrix.
. d · h ‘ activities under Mark the next columns with umes, an write eac person s ed
the appropriate time. Exhibit 12.2 illustrates an example of an integrat
accident event matrix. ft shows that there was no interaction ben~een the
. U d at the p1pefiner pipefitter and the mechanic and that the supervisor ye e .
to get the job done as soon as possible, interrupting him. Creaong an
Chapter 12: SperialiZJd and Co111puterized Tedmiqtm
. . a simple technique that can be used to
ed accident event macnx is
ji11ef1′ . ns between people.
dis(O”er wceracno
odes and Effects Analysis .
failure M al . . sually used as a proacave safety tool d effects an ys1s ts u
,,_:1,,re rnodes an . al al how failures can affect systems. The
r,.,i- – r rofesston s an yze . . d .
.,, help safe[) P d actively for accident invesogaaons an 1s ,., also be use re .d
ttChnique can I if the accident was caused by a system failure. In acc1 ent
rs~ y.usefu all failure modes should be analyzed to determine whether
J11rtsogat1ons,
Exhibit 12.2 C: INTEGRATED ACCIDENT EVENT MATRIX
Note: This technique is very helpful in fata l acciden~s
or other situations where some viewpoints are not available.
PEOPLE 10:01 10:02 10:03 10:04
Usf all people Record what
whocouldgive each worker
infomlotion was doing at
aboutinter- each time in
ocrionsand the sequence.
lttlp toanalyze
whatoc:curred
otthetimeof
rht ocddent.
Pipefitter Walked to job Turned to talk Went to Walked to
site with super- electrical box equipment;
visor and turned on found
breaker mechanic in
pain
Supervisor Talked with Yelled at Finished Finished
vendor pipefitter paperwork for paperwork
pipefitter
Mechanic Worked on Worked on Talked on Received equipment equipment cell phone electrical
– shock
Part TT!: A 11a!Jtiral Terh11iq11es
132
Other Specialized Techniques
Expert Techniques
Some techniques that can be useful in accident investiga,-;
. uOnare b by e.,–pe.rts. E.,;:amples include: est Perfottned
Sofa,,ar, ha-:z.ard ana!Jrsis. Helps investigators to analyze f
so tware fail find causal fuctors in computer systems. llres ‘-nd
Common Cal/St faihm ana!Jsis. Used to find system failures th
. . . at led to accide
• Sneak amal ana!Jn.s. Looks at sneaks (failures) in a s nts.
ystem or circuit.
Failure Analysis and Structural Analysis
Failure analysis and structural analysis can be used for aca·d · .
ent mvesuga · ro determine types of failures and structural flaws that led t . llons
o an acadent.
Sdentific Modeling
]Vlany types of modeling can be used to describe possible accident scenarios or
ro sample data to reconstruct scenarios. These types of analysis are performed
by experts who understand data collection and analysis processes. Most of
these types of analysis are performed in a laboratory setting.
Acddent Reconstruction
Accident reconstructions may be simulated or computer-generated. To
reconstruct an accident, the investigator finds out how each step in the accident
sequence occurred. Once this sequence is determined, the reconstruction
will help ro analyze the effects or potential effects of the each event in the
sequence.
Computerized Techniques
Graphical Programs
. . Mi soft VisioTM and
lnvesogators can use graphical programs such as cro h alytical
h rts grap an CorelDRAWTM ro create events and causal factors c a ‘ d effects
d ture cause an trees, draw barrier analysis summary charts, an strUC
Chapter 12: Specialized a11d Computerized Tech11iq11es
s such as Microsoft PowerPoint™ and Corel ration program . .
, .. ….ens-Presen d t draw simple charts and analyocal trees. While
ill’5′- rM can be use o . . . .
sent:1cions . all di play the results of the accident mvesogaaon, p~ ms graphic y s . .
tJ,
d not actu Y tb
. al Programs
,\Jlalyuc t ized programs however, can help accident b ed of compu er , .
A new re . blem solving, investigation, and analysis. These programs
·”‘tors with pro . . b al ~res0o·· nl find causal facrors and correcave acoons, ut so to be used not o y to
c,J\ al ses into written reports.
– ~m y . .
rograms have been designed for accident analysis,
Many custom p . 1 . U .
d 1 factors charting and analysis, and tree ana ys1s. smg – m= . .. 1· am does not take the place of mvesogaong, ana yzmg, ,computer progr . . .
thi ki however. Four accident invesogaoon and problem solvmg ,nd n ng,
programs are:
, REASON® Root Cause Analysis by Decision Systems, Inc.
, Apollo Root Cause Analysis and RealityCharting™ by Apollo Associated
Services
, TapRooT® System with SnapCharT® Software and Root Cause Tree®
Software by System Improvements, Inc.
‘ RootCause LEADER™ Software by ABS Consulting, Inc.
The following short descriptions of the capabilities of each program were
obtained from the companies’ web sites. Web site addresses are listed in the
bibliography at the end of the book.
REASON® Root Cause Analysis
The latest version of this root cause analysis software from Decision Systems,
Inc. is called REASON@ 7.2. REASON® believes root cause analysis should
be a validated and consistent process that discovers the root cause of a
Ptoblem so th b · – f h at usiness pracoces can be used to prevent recurrence o t e
(‘
0
_blem. One of the key features of REASON® is that it gives the user a
thrailroad track to get you to the correct root cause.” In other words, it gives
e user th l
e too s to find the failure. The program leads the user to ask the
133
Parl III: A 11t1()’hcal Terlmiqu,s
134
right questions, and thus arrive at the e d . n pomt (
an e.,perc sys tem software tha t guides yo root cause) “t> . u to uncov · ,,ea
your operaoons problems, enable you to er the r00 50n is . manage and t cau
acoon plans and communicates the lesson learned track Your c ses of
activities” (Decision Systems Inc. 2011) from Y0 ur probl orrective ernso[ ·
111.is software is used not only for accide · . V!ng . nt mvest::tgations
of problem so!V1.ng or process improvement. A vali . , but for an
c…. ti · · . dat::ton ste . Y type son ,vare; us gives consistency to the root ca al . Pis built in
d
use an ys1s pro to the
oes nor attempt to funnel the user into a sel cess. REAsoN
. . ect number of root ca r t
Other unportant funcnons are a correctiv . Uses. . . e act::ton datab
correcove acoons and a searchable query for hin ase that track searc g previ s
issues, root causes, and corrective actions. The ftw ous accidents so are also h ‘
report editor. This program can be used for both . as a powerful
. . . . reacove and .
s1tuanons and combrnes rntegrated root cause anal . _proacuve
. ys1s, corrective ac .
tracking, and lessons learned. (Decision Systems Inc. 20l l) Uon
Apollo Root Cause Analysis
When Apollo Associated Services first started out it introduc d . . ‘ e an approach
to basic problem solvrng that also works well for accident investigation. A llo
has developed a very effective training program and cause-and-effect pr!:ss.
The cause-and-effect process can be performed either on a computer or
on paper. The process and a particular method of analysis is Apollo’s real
product. As the computer has grown to be a more important tool in accident
investigation, Apollo has developed a variety of computerized charting
programs. The newest is RealityCharting™.
RealityCharting™ is a graphical program that facilitates the cause-and·
effect charting process. One of its features is the ability to drag and drop
causes to any location on the chart. The software also provides:
• embedded Apollo Root Cause Analysis methodology
• detailed problem definition helps you understand sigiuficance
• graphic representation of interrelated causes and causal paths
• effective solution generation
• comprehensive reporting features
-I
S
. ,. ed and Co/Jlputerized Techniques
Chapter t 2: pee1a,1z
. nal user through the rules of the
,,,,ides the new or occas10 Jetion of a RealityChart (Apollo
0ndoW b- th rn roward comp if!I’ d leads e
,\ ~tz ethod llfl ,,,,uo m . s 2011)-~, . red service
-~-o0a . .
o0’f® . two computerized accident invesngat1on
f1pB T®Systern consists of T e® These products are
Roo d R ot Cause re · d
,i,, fap SnapChar’f® an o . dhesive notes to chart events an
grams, t over using a h d pro irnprovernen . al gram that draws c arts an
Jefini1ely an Char’f® is a graphic pro . li
., factors. Snap .fi . The result is a presentation-qua ty
~us~ modi 1cat1ons. d h
. ,.,ms and allows easy . h I investigators understan w at
,lii5•- Ch ‘f® diagram e ps “The Snap ar d h ,, (Systems Improvement
ch•rt- 1 . what happene to ot ers. happened and exp am
J 2011) ff
nc. ot Cause Tree® software picks up where SnapCharT® leaves o • ‘In~:~® detemunes what happened, and Root Cause Tree® finds root
lnap ard develops corrective actions. 111.is software features a Root Cause
~m C Tree®diccionary and a Corrective Action Helper® module. The Root ause
Tre~includes a Human Performance Troubleshooting Guide that helps
investigators ask the right questions to solve human performance problems.
A built-in reporting feature and integrated databases ensure that corrective
,ctions are tracked (Systems Improvement Inc. 2011 ).
RaotCause LEADER™ Software
RootCause LEADER™ allows accident investigators to investigate and track
my l}pe of incident, event, or nushap. It can also perform data trending and
ana!ys15, generate report forms, and include a detailed background/ description
for each causal fact d >-rL: ftw . . . . or an root cause. , ms so are can 1dent1fy root causes of
madents events · d • . . .
b
. • , acc1 ents, near nusses, reliability problems, quality impacts,
or Ustness losses.
RootCause LEADER™ h . . . usino R C as five key features: identifying consequences
-.,, OOt ause Ma TM hin ‘
recornrn d . P ‘ attac g photo files and other documents, tracking
en auons and tr d. Th Roo1ca M ‘ en ing. e consequence categories and ABS’s use apTM ar £ £
investigator identi e eatures or customizing the database and helping the
fy root causes (ABS Consulting Inc. 2011 ) .
135
Part [] I: A nalytical T echnique,
136
Many specialized and
. . cornputeriz d
Summary
accident 1nvestigatio I e anaJyticaJ
h . n. n Order to technj
tee 111que, it rnust be c receive th que,
1 . y y an expe ,, S!tuatton ,, on fro u sed for accident analysis b . rt. <>1any syste~ . “‘any re,, .. ~•
. . , ut cauaon h … safety “‘ll’!lrti gamed 1s useful. These anaJys . s 0 uJd be used t ana.Jys,
8
, , _
th es rnay identify O ensu, ..,. be at rnay not be practical for th .d . an overwhel-, e that th, ,. al . e ace, ent tn . “‘ung “‘ll an ys1s and problem-so[v;~g p vesttgati
00
_.,,
. u, rograrns b •nenew to tnvestigators. are ecorning i . c
1. What does time loss analysis try to analyze?
2. Which techniques must be performed b
ncreastngl , .
l rn,Po”‘n,
y an expert?
3. What does a failure modes and effects analysis look for?
4. What is the difference between a design criteria analysisand h
c angeanalysisl
5. How can computerized techniques be helpful in accident investigations?
Part IV
< z
e
PREVENTING ACCIDENTS
D · d hich is the ultimate f, how to prevent future ace, ents, w . art: :;~:~::cnting accident investigations. Determining the accident
purpo d the causal factors prepares the accident investigator to deterrrune
stljUence an · · hould
corrective actions that will prevent similar accidents. Correcuve acuons s
be initiated, documented, and followed up (audited) to ensure that they are
perfonning as intended.
Writing an accident report or filling out an accident form is not simply an
exercise in paperwork. The report should document that the facts and analysis
are correct, the accident sequence has been determined, and corrective actions
have been developed to avoid recurrence of the accident.
Learning from accidents is one of the important aspects for conducting
an accident investigation. Communicating the lessons learned, preventing
systemic problems, and improving the safety programs are key components
of learning from the accident. No accident should be repeated.
Objectives for Part IV:
U
nd
er
st
and the purpose of corrective actions and be able to develop
effecuve corrective actions and recommendations for accidents.
Understand th I · hi
. e re auons ps among facts, analysis causes and correcu·ve , ,
